Two randomized-block experiments were carried out to examine the relative value of wheat and barley as supplements to grass silage for finishing beef cattle. In each experiment unwilted, formic acid-treated silage was offered ad libitum and supplemented with 500 g soya-bean meal and 50 g minerals and vitamins to 44 12-month-old bulls for 157 and 172 days in experiments 1 and 2 respectively. Twelve of the animals also received 2·5 kg rolled spring barley (LB), 12 received 4·0 kg barley (HB) and 20 received 3·25 kg rolled wheat (W). For experiments 1 and 2 respectively the barley contained 796 and 787 g dry matter (DM) per kg; 118 and 105 g crude protein (CP) per kg DM; 47 and 57 g crude fibre per kg DM; the wheat contained 845 and 800 g DM per kg; 112 and 116 g CP per kg DM; 23 and 25 g crude fibre per kg DM; and the silages contained 190 and 177 g DM per kg; 153 and 176 g CP per kg DM; 80 and 104 g ammonia-nitrogen per kg total nitrogen. On average over the two experiments, for treatments LB, HB and W respectively, silage DM intakes were 5·4, 4·7 (s.e. 0·14) and 4·9 (s.e. 0·11) kg/day; total DM intakes 7·9, 8·3 (s.e. 0·14) and 8·1 (s.e. 0·11) kg/day; metabolizable energy intakes 91·4, 97·8 and 94·2 MJ/day; live-weight gains 1·04,1·19 (s.e. 0·029) and 1·10 (s.e. 0·023) kg/day and carcass gams 0·65, 0·77 (s.e. 0·017) and 0·70 (s.e. 0·013) kg/day. It is concluded that the feeding value of wheat was proportionately 0·98 of that of barley for finishing beef cattle when given as a supplement to grass silage, and that the type of cereal offered did not affect silage intake or carcass composition.

A total of 161 bulls from 11 breeds were used to generate crossbred calves from Angus cows in 1973 to 1977 at each of three diverse New Zealand locations, and from Hereford cows at one of the locations in the same years. The bulls comprised four local breeds, Angus, Friesian, Hereford, and Jersey, and seven recently imported breeds, Blonde d'Aquitaine, Charolais, Chianina, Limousin, Maine Anjou, Simmental and South Devon. This paper reports the reproductive and maternal performance of the straightbred and first-cross cows over the first four calvings, with first mating as yearlings at 14 to 16 months of age. A total of 7575 mating records from 2109 cows were analysed. Location differences were greater for reproduction than for growth traits and resulted in genotype × environment interactions for some components of cow performance and particularly the composite traits, weight of calf weaned per cow joined (productivity) and cow ‘efficiency’ (the ratio of productivity to cow weight). At all locations the Friesian-cross cows weaned the greatest weight of calf per head but were matched or surpassed by the lighter Jersey crosses in terms of efficiency of calf production. Most of the European crosses performed relatively much better in the most favourable environment than in the harsh environment and this was particularly marked for the productivity of Simmental crosses. Heterosis as a proportion of the purebred mean was important for cow performance and particularly pregnancy rate (0·12), productivity (0·21) and the efficiency ratio (0·16). Heritabilities for weight and age at puberty were both 0·34 (s.e. 0·08). Repeatabilities and heritabilities for cow reproductive traits were low (0·0 to 0·10) but higher for calf weights up to weaning as a trait of the cow (0·09 to 0·38). In general, the large European breeds which excelled in growth and carcass production produced female progeny which reached puberty at greater ages, had lower reproductive performance (especially in less favourable environments) and larger mature size. Some breed utilization strategies to achieve trade-offs between these genetic antagonisms are discussed.

Eighty castrated male lambs were allocated to either an initial slaughter group (no. = 16) or one of eight treatment groups (no. = 8). Ammoniated barley straw was offered ad libitum. This was supplemented with sodium (Na) propionate (0 or 20 g/day), formaldehyde-treated casein, (protected casein, 0 or 50 g/day) and/or calcium (Ca) salts of long-chain fatty acids (Ca soap, 0 or 30 g/day), as a factorial design. Measurements were made of food intake, live-weight gain and wool growth, rumen fluid parameters and changes in body composition over a period of 155 days.

Food intake (g dry matter per kg live weight) was lower in the groups supplemented with Ca soap and Na propionate. Fleece-free live-weight gain was increased by supplementation with protected casein and Ca soap, but was not affected by Na propionate. Supplementation with protected casein increased wool yield. Adjusted means for water, protein and ash content of the body (adjusted to mean fleece/digesta-free body weight at slaughter of 27·1 kg) were affected significantly by the second-order interaction between the supplements. Adjusted body water and protein mass were lower when Na propionate was given alone or with Ca soap and protected casein. Fat mass tended to be higher under those conditions. Supplementation with Ca soap increased adjusted fat and gross energy contents and decreased adjusted protein and water mass. Supplementation with protected casein decreased total body fat and increased adjusted protein mass in the body. The partitioning of water, protein, fat, ash and energy between the carcass and non-carcass residues was affected to a small extent also. These results demonstrate interactive effects of the supplements used, on nutrient partitioning in lambs offered straw-based diets.

An experiment was designed to determine whether combining a single injection of 125 μg prostaglandin F2α analogue (PGF) with a 5-day progestagen pessary treatment (PPT) would obviate the problem with refractory animals that occurs when randomly cycling ewes are synchronized with PGF alone. The efficacy of the combined treatment, in comparison with single and double dose PGF regimes, was also assessed.

A total of 210 ewes were used in this investigation conducted during the 1989 and 1990 breeding seasons. In both breeding seasons, 100% of the ewes treated with PGF combined with PPT were mated by fertile rams within a 3-day period, compared with 52·9% of ewes treated with a single injection PGF or 100% of ewes treated with two injections of PGF, 11-days apart. Returns to service for the three groups were 14·2%, 16·2% and 17·4%, respectively. There was no effect of treatment on lambing performance in 1989. However, in 1990 there were differences between groups in gestation length and birth weight, but these were attributable to enzootic abortion. The results demonstrate that PGF combined with PPT will produce levels of oestrous synchronization and fertility comparable with other methods.

The effect of including 14 g urea with either 75 g dry matter (DM) from sugar cane molasses (UM) or Colombian rice polishings (RP) at three levels, 68 (RP1), 137 (RP2) or 203 (RP3) g DM on grass hay DM intake and on rumen fermentation was investigated. An incomplete Latin-square design was used and each experimental period was divided into 12 days for adaptation to each diet followed by 9 days in metabolism cages when all measurements were made using five adult sheep.

Pooled mean values for rumen metabolites (five samples per day) were calculated. Rumen pH was not affected by the nature of the supplements. Concentrations of volatile fatty acids (VFA) (UM 92·8, RP1 84·2, RP2 86·4, and RP3 84·0 (s.e.d. 3·4) mmol/l) and lactate (UM 2·0, RP1 1/6, RP2 1/7, RP3 1/8 (s.e.d. 0·014) mmol/l) (UM v. RP, P < 0·05 and P < 0·10 respectively) were lower when RP were given, while concentration of branched and longer chain VFA (26·5, 34·0, 31·1 and 33·5 (s.e.d. 1·6) mmol/mol total VFA, UM v. RP, P < 0·01) and ammonia (98, 131, 141, 137 (s.e.d. 16·1) mg/l, UM v. RP, P < 0·05) were increased. Numbers of rumen protozoa (1·6, 3·2, 2·7, 3·3 (s.e.d. 0·75) × 105 per ml, UM v. RP, P < 0·20) tended to be higher 2 h after feeding when RP rather than UM were given. However, hay DM intake (1050, 960, 960, 880 (s.e.d. 45·3) g/day, UM v. RP, P < 0·05), and microbial protein supply (11·7, 9·3,11·1,10·8 (s.e.d. 0·59) g N per day, UM v. RP, P < 0·05), estimated from urinary purine excretion were reduced by feeding RP instead of UM.

At the levels of inclusion tested RP did not increase the efficiency of rumen fermentation and were not as effective a supplement with urea as was molasses for a low-quality forage diet.

The objective of this work was to determine the apparent bioavailability of magnesium from four different types of calcined magnesites of Greek origin. These four types (A, B, C, D) were different in particle size fractions and production process. The solubility of the above four magnesium samples and their fractions were determined: (a) in the laboratory using ammonium nitrate solution after 24-h shaking and (b) in vivo using a nylon bag technique after 24-h incubation.

Solubility was correlated with apparent bioavailability as determined in a balance experiment with five wether sheep in a 5×5 Latin-square design.

Sheep were kept in metabolism cages and were given daily a basal diet (500 g hay + 700 g concentrate) plus 2 g magnesium supplements from types A to D. The magnesium bioavailability of the basal diet was used as a base for the determination of the bioavailability of magnesium in A, B, C and D supplements.

Magnesium apparent bioavailability of the basal diet was found to be 0·203 and, for the supplements A, B, C and D, 0·381, 0·339, 0·315 and 0·292 respectively. The solubility of supplements A, B, C and D in ammonium nitrate solution and in vivo (using the nylon-bag technique) was found to be 0·812, 0·805, 0·784 and 0·773 or 0·294, 0·152, 0·102 and 0·200 respectively.

The solubilities of the magnesium supplements in ammonium nitrate solution and in vivo both reflected dietary bioavailability differences up to a point, but neither was well enough correlated with the apparent bioavailability as determined by the balance experiment, this being judged to be the most precise method for bioavailability determination.

Telemetric techniques for the monitoring of physiological parameters during housing, handling and transport may help producers to reduce mortality and improve meat quality. In order to know the reliability of electrocardiogram parameters as stress indicators, piglets (12 to 23 kg) being different with respect to the halothane gene (homozygous halothane sensitive (nn) and non-sensitive (NN), heterozygotes (nN)), and thus different with respect to stress susceptibility, were monitored with an ambulatory ECG device during housing, handling and transport. Skeletal muscularity of all animals was measured with an ultrasound device.

Heterozygotes and homozygote halothane sensitive piglets had a higher muscularity, but only the homozygotes had a higher ST-elevation from the isoelectric line of the QRST-complex of the electrocardiogram when measured within housing conditions. This means a different propagation of the electric current related to the activation of the ventricle. Halothane gene carriers (Nn and nn piglets) had a statistically significant higher mean maximal heart rate during housing conditions. During handling heart rate rose by 14·8%, 0·61% and 42·9% respectively for NN, nN and nn piglets, with nn piglets having the highest values. During transport these measurements were respectively 6·88%, 3·66% and 9·7% higher for lines NN, nN and nn as compared with those in housing conditions. Heart arrhythmicity parameters were not different between housing and transport, but during handling up to a 23-fold increase could be observed.

Heart rate and arrhythmicity parameters may generate reliable information to monitor stress during housing, handling and transport of piglets through genotypes and skeletal muscularity.

Sixty castrated male lambs fca. 37 kg) were allocated to an initial slaughter group (no. = 12) or to one of six treatment groups (no. = 8). Chopped oaten hay (sprayed with 10 g urea per kg) was offered ad libitum with 57 gjday chopped lucerne hay. Additional supplements were 0 or 57 g/day formaldehyde-treated casein (protected casein) and volatile fatty acids (VFAs) at ca. 1·45 MJ gross energy per day. The VFAs given were acetate or propionate or a mixture of these (molar ratio 4: 1). Daily food intake and weekly live-weight (LW) gain were measured and rumen fluid and blood were collected. The lambs were shorn, slaughtered and body composition was determined.

Food intake (g/kg LW) was not affected by treatments. Supplementation with protected casein increased LW gain and wool growth. Supplementation with propionate reduced LW gain, but not when given with protected casein. Plasma concentrations of urea-nitrogen and insulin were increased, and plasma somatotropin decreased in lambs given protected casein. Final body content of water, fat, protein and gross energy, adjusted to the mean fleece/digesta-free body weight of 36·5 kg, was not affected by the treatments. The partitioning of water, fat and energy between the carcass and the rest of the body was affected by the treatments. Excretion of VFAs in urine was measured in two sheep in experiment 2, and was negligible compared with the quantity of VFAs given. It is concluded that energy from salts of VFAs, in particular propionate, is used inefficiently for body-weight gain in lambs given low-protein roughage diets. The nutritional treatments had only marginal effects on the composition of body gain.

The effect of proportion of dietary rumen non-degradable protein during food restriction on the weight of components of the digestive tract, carcass and subsequent growth was investigated in 37 Scottish Blackface wether lambs (initial live weight, 25 kg). Lambs were given individually ad libitum for 4 weeks a complete pelleted diet containing 150 g/kg white fish-meal (HP, 10·4 MJ metabolizable energy (ME) per kg dry matter (DM), 195 g crude protein (CP) per kg DM). Seven lambs were then slaughtered and 15 were switched to a diet containing no white fish-meal (LP, 10·4 MJ ME per kg DM, 122 g CP per kg DM) offered at 18 g DM per kg M per day, i.e. sufficient to maintain constant live weight. A further 15 continued to receive the HP diet at the same rate as the lambs given LP. After 6 and 12 weeks, five lambs on each diet were slaughtered. At 12 weeks the remainder received the HP diet ad libitum for a further 7 weeks before slaughter. During food restriction on both diets, the proportion of live weight formed by the carcass, the dissected components and the chemical composition remained the same as in the initial slaughter group. The relative weight of the non-carcass components fell during food restriction on both diets. There was a significant (P < 0·05) interaction between slaughter date and dietary treatment for reticulo-rumen weight as a proportion of empty body weight (EBM); it was smaller in lambs on the HP diet after 12 weeks of restriction (HP: 0·022, LP: 0·026). A similar pattern was observed for the small intestine and the total digestive tract. During the 7 weeks of realimentation, lambs previously on HP and LP diets had similar intakes (1343 and 1208 g DM per day) and digestive tract components, body components and chemical composition of carcass and non-carcass components were all unaffected by previous treatment. The apparent digestibility of CP of the HP diet was greater than that of the LP diet although it was overall less degradable in the rumen. When both groups of lambs were realimented on the HP diet, there were no differences in the apparent digestibility of any of the dietary components. A high dietary protein: energy ratio during restriction reduced the weight of some of the components of the digestive tract but did not significantly affect carcass composition. The effect did not persist following realimentation and did not significantly influence subsequent performance.

Heat production, energy and nitrogen balances of artificially reared piglets from specific pathogen free sows, housed in a respiratory chamber, were measured over five successive periods of 7 days each from 4 to 39 days of age. The experiment was carried out at an environmental temperature decreasing from 33 to 27°C. The mean weight increased from 1·5 (at day 4) to 11·9 kg (at day 39). Daily rate of gain increased from about 100 g in periods 1 and 2 to 502 g per piglet in the fifth period.

Daily protein deposition, as measured via nitrogen balance, increased from 17 to 96 g; fat deposition increased from about 0 to 100 g. Body composition at the start and end was determined by a comparative slaughter technique. Protein and fat content had a high correlation coefficient with body weight (R2 = 0·69 and 0·89, respectively). At birth these values were much lower, 0·24 and 0·01 for protein and fat, respectively.

Nitrogen balance compared with slaughter technique led to an overestimation for protein deposition of proportionately 0·07. Fat gain measured from slaughter was proportionately 0·08 lower than that calculated from energy balance.

Ratio of gain estimated from protein and fat gain with the balance technique relative to that measured ranged between 0·93 to 1·15 in various periods.

Data on 3783 Yorkshire, 2842 Landrace and 937 Duroc littermate pairs of boars, station tested between 1976 and 1989, were used to measure genetic parameters of average daily gain (ADG) from 30 to 90 kg, live backfat at 90 kg (BF), estimated lean growth rate (LGR), average daily food intake, food conversion ratio (FCR) and lean food conversion ratio (LFCR), as well as measures of residual daily food intakes over and above requirements for growth and lean growth. A method was developed to obtain restricted maximum likelihood estimates of genetic variances and covariances under an animal model when observations are on the means of sib pairs. Heritabilities of ADG, BF, LGR, FCR and LFCR were 0·43, 0·59, 0·39, 0·28 and 0·34, respectively. Heritability of daily food intake was 0·45, and heritability of measures of residual daily food intake ranged from 0·30 to 0·38. About half of the variation in daily food intake was residual (0·48 to 0·56). Genetic correlations of ADG with daily food intake, FCR and LFCR were 0·80, −0·28 and −0·09, respectively, and were small and positive (0·18 to 0·34) with measures of residual daily food intake. Backfat had genetic correlations of 0·42, 0·24 and 0·52 with daily food intake, FCR and LFCR, respectively, and genetic correlations between backfat and measures of residual daily food intake ranged from 0·15 to 0·61. Selection against residual food intake may be a useful means of improving efficiency of food utilization.

Thirty-six wether lambs were given food hourly consisting of mature chopped oaten hay, sprayed with 10 g urea per kg hay. A pelleted supplement was given containing 0 or 75 g/day fish meal and 0 or ca. 1·5 MJ/day gross energy as acetate or propionate in a 2 × 3 factorial design. Dry matter (DM) digestibility, nitrogen retention and urinary allantoin excretion were determined over 7 days. Tissue protein kinetics were estimated using 15N-glycine. Net fluxes of acetate, bicarbonate and glucose were estimated with continuous infusion of 14C-acetate, 14C-bicarbonate and 3H-glucose respectively.

Intake of oaten hay (g DM per kg live weight per day) was reduced by supplementation. Live-weight gain was lower in lambs supplemented with propionate than in lambs given the basal diet only. Supplementation with fish meal increased DM digestibility, live-weight gain and wool growth. Daily excretion of urinary allantoin did not differ between groups of lambs. Nitrogen retention and protein flux were reduced in lambs supplemented with acetate or propionate and were increased in lambs given fish meal. Plasma acetate concentrations were higher for lambs supplemented with acetate. However, net flux of acetate did not differ between groups of lambs. The net flux of glucose was increased by supplementation with propionate and with fish meal. Plasma levels of β-hydroxybutyrate were reduced in lambs supplemented with propionate.

Availability of propionate and glucose were not limiting live-weight gain in these lambs. Also, a high ratio of acetate to propionate did not reduce the apparent efficiency of nutrient use.

The interpretation of the results of feeding trials with lactating ewes and their sucking lambs can be improved considerably when milk energy production is known. The determination of gross energy (GE) however is time consuming and expensive. Therefore an equation to predict GE from milk constituents would be helpful.

Using multiple regression analysis a GE-prediction equation was derived with milk samples mainly from crossbred ewes. The concentrations of milk constituents were determined by infrared spectrometry (calibrated with cow milk). Fat concentration ranged between 43·8 and 125·0 g/kg, protein concentration between 32·4 and 53·1 g/kg and lactose concentration between 38·9 and 52·9 g/kg. GE (adiabatic bomb calorimetry) of the samples under consideration ranged between 3500 to 6800 kj/kg.

The following equation, including fat (f), protein (p) and lactose (I), is recommended and predicts GE as kj/kg fresh milk using constituents in g/kg fresh milk:

GE = 41·94 × f + 15·85 × p + 21·41 ×l (residual s.d. = 92, adj.R2 = 0·98).

This equation has, in the GE range 4500 kj to 6000 kj, an almost constant confidence interval with an average of ±25 kJ and an almost constant prediction interval, with an average of ±87 kJ. In the period until weaning it would appear justified to use the derived equation regardless of stage of lactation. The equation is valid for milk samples from sheep breeds with relatively high fat and low protein contents and where milk constituents have been determined by cow milk calibrated infrared spectrometry.

Predicting the performance of animals is a general problem in animal production, the solution to the problem depending, in part, on being able to describe the animals adequately. Numerous approaches have been made to model the growth of pigs, with different methods being employed to describe the genotype. A simple approach is to describe the potential rate of growth of body protein using a Gompertz equation, and to predict the growth of the other chemical components of the body using allometry. These components are added to determine the body weight of the animal.

In this paper an experimental procedure is proposed that will predict values for the parameters of the Gompertz growth equation under non-limiting conditions. These parameters include the rate of maturing of the animal and the mature protein weight. The lipid-to-protein ratio at maturity is required to describe the allometry between protein and lipid. With these few parameter estimates the potential growth rate of animals can be described. It is then possible to use this information to predict the growth rate of animals under conditions constrained by the food or the environment.

Scottish Blackface sheep have been selected for either increased or decreased predicted carcass lean content at 20 weeks of age, using an index combining ultrasonic backfat depth and body weight. The index was designed to alter carcass lean content without changing body weight. After 4 years there have been large responses in the index and fat depth, the proportional difference between the divergent lines in fat depth is 0·28, but body weight and ultrasonic muscle depth have not changed between the lines. The realized heritability for the index is 0·45. Heritabilities for the index, ultrasonic fat depth, body weight and ultrasonic muscle depth, calculated using multivariate restricted maximum likelihood, are 0·47 (s.e. = 0·14), 0·39 (s.e. 0·13), 0·23 (s.e. 0·12) and 0·36 (s.e. 0·14), respectively. The selected lines will be used to investigate relationships between carcass lean content, viability and maternal performance under the harsh environmental conditions encountered in the Scottish hill environment.

Previous studies in foetal and growing sheep have shown that prolonged administration of β-agonist drugs leads to selective but marked attenuation of cardiovascular, metabolic and endocrine responses to the natural catecholamine hormones adrenaline and noradrenaline, as well as attenuation of responses to the drug itself. The experiment reported here was carried out to determine whether administration of the β-agonist drug, salbutamol, by twice-daily intramuscular injection at rates of 40 or 200 μ/kg per day, for a period long enough to result in substantial attenuation of its metabolic and endocrine effects, might also result in impairment in the ability of treated animals to respond to stressful husbandry procedures which disturb metabolic homeostasis and which depend on increased sympathetic nervous and adrenal activity for the restoration of homeostasis. Serial blood sampling on day 1 and day 14 of salbutamol treatment showed that the large increases in plasma lactate and glucose observed on the 1st day were absent on day 14, while the rapid increases in free fatty acids and insulin observed on day 1 were very greatly attenuated. Daily blood sampling also demonstrated that salbutamol, like cimaterol, significantly decreased both pre- and post-feeding plasma insulin and glucose concentrations.

Metabolic and endocrine changes consequent on herding and immersion of the lambs in a sheep-dip, were unaltered by 14 days of salbutamol treatment, even though the procedure resulted in activation of the pituitary-adrenal axis as evidenced by increased β-endorphin and cortisol concentrations, and a large increase in plasma lactate concentration. These results suggest β-agonist-treated animals can respond normally to physiological stresses but further investigations remain necessary to determine whether responses dependent on increased lipid mobilization and shivering, such as shearing or prolonged exposure to severe cold, remain normal in β-agonist treated animals.

An experiment was designed to quantify the specific genetic parameters which will describe the growth potential of a pig. The experiment was conducted in two stages. In the early stage, 11 unimproved crossbred pigs were serially slaughtered at 2, 13 and 23 days of age. In the later phase, 16 pigs were serially slaughtered at 80, 90 and 100 kg live weight. The chemical composition of each of the entire empty bodies was determined. Linear regression analyses were performed on body protein and lipid weights to determine the Gompertz growth parameter (B), mature body protein weight (Pm) and the lipid: protein ratio at maturity (LPRm) The B value for boars was 0-0107 (s.e. 0·00060) per day and for females 0·0120 (s.e. 0·00044) per day. Pm estimates were 38·7 kg and 28·4 kg whilst the LPRm ratios were 2·60 and 3·89 for boars and gilts respectively.

When designing environments for animals, the quality of space as well as the amount of space should be considered. In this study the influence of an opaque barrier on the aggressive behaviour and growth of Large White Χ Landrace piglets that were weaned at 24 days was examined. The behaviour and weight gain of 120 piglets were recorded during a 4-week period following weaning. These piglets were housed in either a conventional flat-deck pen or a straw pen, with or without an opaque barrier. There was no significant difference between the two types of accommodation in the frequency of aggressive interactions that occurred during the weaning period but where a barrier was present the frequency of aggressive interactions was 40% lower during the 1st week after weaning and growth rate was proportionately 0·15 greater. Piglets that were the recipients of most aggressive behaviour used the barrier most frequently during the weaning period and barrier users gained the most weight during the week immediately following weaning.

It was concluded that a barrier improved the weaning environment because it offered piglets an escape route during the period when most aggressive interactions occurred. A barrier may be an effective way of improving the welfare of animals housed in confined conditions.

Amine-organic buffers (BESKOH and TESTRIS) were compared for their ability to be used for buck semen dilution in a freezing procedure. BESKOH showed better results (P < 0·05) at the time of dilution and after chilling at +5°C (72% motility v. 54% at dilution and 62% v. 39% after chilling). After freezing and thawing, none of the variables measured (motility, normal acrosome ridge and normal swelling) was significantly different. A washing procedure was necessary when freezing was undertaken and egg yolk addition was necessary to avoid cold shock as the temperature lowered during chilling. There were no significant differences between a 3% and 12% level of egg yolk addition.

Responses of gross energy intake and body weight of mice from weaning to maturity to dietary composition of energy-yielding nutrients were studied. Twenty-four groups of ddY male mice were fed for 70 days purified diets with a range of 0·08 to 0·75 protein, 0 to 0·92 fat and 0 to 0·92 carbohydrate concentration, calculated on a gross energy basis. Food intake and body weight data in each group were analysed by non-linear regression to obtain values of the parameters in the feeding and growth equation of Parks (1982). Prediction equations for these parameters were obtained with mixture model forms as a function of dietary protein, fat and carbohydrate. These equations for the parameters were then used to construct the response surfaces of body weight and growth rate at a given time on a triangular graph. The response surfaces of body weight at any time were convex, peak positions of which shifted from 043 protein and 0·29 fat at 10 days to 0·36 protein and 0·52 fat at 70 days. The configuration of response surfaces of growth rate (dW/dt) to the three components depends on feeding periods and at the longer periods of feeding the convex configuration of the surfaces became planar. The response surfaces of food efficiency (dW/dF) at any age were almost parallel to those of dW/dt, whereas a parallelism was observed between response surfaces of dW/dt and dF/dt only in young animals. These results suggested that the responses of body weight in young animals to dietary composition of energy-yielding nutrients depended on those of daily energy intake as well as food efficiency, but that those in older animals were affected by food efficiency only.