Three experiments were carried out to obtain more information upon the effects of varying the plane of feeding for bacon pigs.Exps. 1 and 2 followed the same 3 × 2 factorial design in which (a) diets were fed with or without procaine penicillin and (b) a comparison was made between very high (VH-VH), high (H-H) and low (L-L) planes of feeding for pigs from about 8 weeks old to slaughter at about 200 lb. live weight. The planes of feeding were defined according to the amount of t.d.n. offered to each pig daily.For each of these two experiments eight lots of six litter mates were used. One pig from each litter was allocated to each of the six treatments, and although litter mates were housed as a group, all pigs were fed individually in separate compartments.Exp. 3 was a 2 × 3 × 2 factorial to compare two breeds and three planes of feeding during 2 years. During each year three lots of pure-bred Landrace and three lots of Wessex Saddleback × Large White litter mates were used. Each lot consisted of three gilts and three barrows and one pig of each sex was allocated to each of the three feeding treatments. From the start of the experiment, when the pigs were about 8 weeks old, to about 100 lb. weight, all animals were kept to a very high (VH) plane of feeding. From then to slaughter at about 200 lb. live weight the following planes of feeding were compared: (1) very high plane (VH), (2) increasingly restricted plane (R) and (3) low plane (L). As in Exps. 1 and 2 these planes were denned in terms of t.d.n., and litter mates were housed together but fed individually in separate compartments.In both Exps. 1 and 2 pigs with procaine penicillin in their feed grew very slightly faster and more efficiently than those not fed antibiotic, but the differences did not reach the 5% level of statistical significance. The feeding of antibiotio did not affect the response of the pigs to variations in the plane of feeding.In comparison with pigs kept to the VH-VH plane of feeding, pigs kept to the H-H and L-L planes grew 13 and 22% more slowly in Exp. 1, and 11 and 26% more slowly in Exp. 2, but plane of feeding had no effect upon food conversion efficiency in either experiment. These results are for the total experimental period.In both Exps. 1 and 2 reductions in plane of feeding had no significant effect upon length of carcass or thickness of streak, but they caused increases in killing-out percentage and decreases in all measurements of fat thickness. They also caused increases in the area of ‘eye’ muscle (statistically significant in Exp. 2, but not in Exp. 1) and a lengthening in the shape of the ‘eye’ muscle in the bacon rasher (statistically significant in Exp. 1, but not in Exp. 2).In Exp. 3 the Wessex crosses grew faster and more efficiently than the Landrace, but these differences were considerably larger during one year than during the other. In comparison to the Landrace, the Wessex-cross carcasses had a tendency to be shorter (P < 0·10) and to have thicker shoulder fats. They also had 1% more ‘fore’ and 1% less ‘ham’.Compared with pigs kept to the VH-VH plane of feeding, those kept to the VH-R and VH-L planes grew 18 and 36% more slowly, respectively, during the finishing period. Over the total experimental period there was no difference in food conversion efficiency between pigs kept to the VH-VH and VH-R planes, but efficiency was poorer by 4–5% with the VH-L plane. There was no breed × feeding interaction in the data for growth rates or food conversion efficiencies.In Exp. 3 plane of feeding had no significant effect upon killing-out percentage, shape index of ‘eye’ muscle or thickness of streak. The two lower planes of feeding caused reductions in the average rump fat, fat (1) over the ‘eye’ muscle and in the proportion of ‘middle’ in the carcasses; they also caused increases in the proportion of ‘ham’ in the carcasses.Other carcass quality data contained significant interactions which complicated their interpretation. The effects of plane of feeding were inconsistent between:(A) years, for length of carcass, maximum shoulder fat, area of eye muscle and iodine values of the back fat;(B) breeds, for minimum back fat and percentage ‘fore’;(C) sexes, for dressing percentage and fat (2) over ‘eye’ muscle;(D) breeds and years, for dressing percentage;(E) breeds and sexes, for maximum shoulder fat.The results of the three experiments have been compared with other data on planes of feeding from Rowett Research Institute. A table has been given showing the probable average effects of using different planes of feeding in preference to a VH-VH plane for pigs to be slaughtered at about 200 lb. live weight. The variability in these effects under different circumstances has been discussed and a note has been made of the relative improvements in carcass quality which can be affected through feeding and through breeding.