The experimental material consisted of 42 sows divided into 3 feeding groups, each containing 14 sows fed complete compound feed varying in protein content. Protein and amino acid content of the feed fed to the control group was in compliance with Polish standards. Protein content in the experimental groups was reduced by 10% and 20%, respectively. At the same time, lysine, methionine, threonine and tryptophane content was supplemented to the level of the control group. All pregnant and lactating sows were fed individually. The condition of the sows was determined by measuring the thickness of backfat at the P2 position on days 30 and 105 of pregnancy and on day 25 of lactation. The data analyzed in the study included: the total number of piglets born, their body weight gain, milk composition, weaning-to-estrus interval and the farrowing rate. Protein content of the compound feed did not have a significant impact on the increment in backfat during pregnancy and losses during lactation. The sows fed compound feed with reduced protein content gave birth to 0.7 and 0.6 less piglets per litter than the control animals. However, due to lower losses, they bred 0.3 and 0.4 more piglets than the control sows. Average body weight of a piglet on day 21 was 0.5 kg lower in the experimental than in the control group. Dry matter, protein, fat and lactose content of the sow's milk did not depend on protein content of the feed. The weaning-to-estrus interval in the experimental groups was one day longer than in the control group. The reproductive rate accounted for 86% and was comparable in all the groups.