Wheat is a major raw material included, often at high rates, in diets for poultry in many countries throughout the world. Although traditionally regarded as having moderately uniform nutritional value, evidence is accumulating that there is considerable variablility in apparent metabolisable energy (AME) values, particularly eith young birds. Considerable effort has been expended in attemting to explain why this should be so, and to develop laboratory procedures whereby it may be predicted. Much attention has been focused on the non-strach polysaccharide (NSP) fraction. The major constituent of wheat is strach which is consequently the principal energy yielding component. Any factors which influence starch digestibility will therefore exert an impact on AME. Strong positive relationships between starch digestibility and AME have been established. An in vitro system has been developed which demonstrates that extrinsic rather than intrinsic factors are responsible for variations in the digestibility of starch in vitro as it has been shown that straches isolated from wheats with low AME values are hydrolysed to a similar extent in vitro as straches from wheats with high AME values. Differences in starch hydrolysis in wheat meals in vitro (i.e. strach in situ) were observed between high and low AME wheats, suggesting that, following subsequent validation, the system may offer considerable potential as a means of characterising wheat for broilers.