Boar and gilt pigs from three genotypes (with 0, 0·25 and 0·50 Duroc inclusion level) were reared from 30 to 90 kg live weight on seven feeding regimens which involved combinations of diet formulation and feeding level. A diet of conventional energy and protein level (CEP; 14·2 MJ/kg digestible energy, 205 g/kg crude protein, 10 g/kg lysine) and a diet of higher energy and lower protein (HELP; 14·7 M]/kg digestible energy, 166 g/kg crude protein, 7·0 g/kg lysine) were used. One treatment involved feeding the HELP diet ad libitum, with the other six treatments involving feeding the CEP diet ad libitum, restricted, or in combinations of ad libitum and restricted. A total of 721 animals comprising similar numbers of boars and gilts were used to estimate sex differences for growth performance, carcass and meat quality, and eating quality. Tissue growth rates were predicted from regression equations based on P2fat depths or ham-joint dissection, developed from subsamples of animals that were subjected to full-side dissection. Overall when compared with gilts, boars grew faster (838 v. 799 gtday, P < 0·001), had improved food conversion ratios (2·39 v.2-55, P < 0·001) but had similar daily food intakes and lean and subcutaneous fat growth rates. Lean tissue food conversions did not differ significantly between the sexes. Killing-out proportions were higher for gilts (0·766 v. 0·749 , P < 0·001), however P2 backfat thickness and lean proportions did not differ between the sexes. The firmness of the mid-back fat, assessed subjectively and using a penetrometer, was greater for gilts than for boars. The tenderness of pork loin chops, assessed by a trained sensory panel, was judged to be better for boars than for gilts but there was no sex difference in overall acceptability. The fat from boars had a higher level of abnormal odour and boar odour. There was a sex × dietary treatment interaction for boar odour with the HELP diet producing the highest levels and the biggest difference between the sexes for odour scores.