1. A co-ordinated trial has been carried out at eighteen centres to study the effects of feeding various quantities of high-protein supplement to growing-fattening pigs. In treatment 1, 10% white fish meal was included from weaning to 150 lb. live weight and 3% soya-bean meal was included from 150 lb. live weight to slaughter (approximately 210 lb.). Treatment 2 was the same as treatment 1 except that only 7% white fish meal was included during the first stage. Treatment 3 was the same as treatment 2 except that the rations were changed at 1001b. live weight instead of 1501b. Treatment 4 was the same as treatment 3 except that highprotein supplement was omitted altogether after 1501b. live weight; thus there were three rations involved in treatment 4—7% white fish meal from weaning to 100 lb. live weight, 3% soya-bean meal from 100 lb. to 150 lb. and no high-protein supplement thereafter.

2. Growth rate and feed conversion both differed significantly between treatments. Pigs on treatment 1 grew 5·0% faster than those on treatment 4; the average growth rate for treatments 2 and 3, the difference between which was negligible, was 3·3% higher than for treatment 4. Feed conversion for treatment 1 was 5·4% better than for treatment 4 and the average for treatments 2 and 3 was 3·8% better than for treatment 4.

3. Treatment differences were not significant for killing-out percentage, length, quality of fat, proportions of gammon, middle and fore-end of the cured side, thickness of streak, size of eye muscle and amount of fat over it. Some of the fat measurements differed significantly between treatments; differences in fat thickness at the shoulder, middle and rump between treatments 1, 2 and 3 were generally small, but the lowest protein level (treatment 4) resulted in carcasses with about 2% more fat along the back than the average of treatments 1, 2 and 3. These slight differences in back-fat thickness did not result in significant differences in the percentage of pigs in the highest grades.

4. An economic appraisal of the results suggested that under the conditions prevailing at the time of the experiment treatment 3 would have resulted in the greatest profit per year, and the greatest profit per pig. This ration was relatively cheap compared with those of treatments 1 and 2 but it did not lead to a seriously impaired performance, as did that of treatment 4, the cheapest ration.