Regression analysis was used to evaluate the effects of gilt age and body composition at first breeding on sow performance over three parities. Eighty-seven Yorkshire x Landrace F1 gilts were used. Variation in age and body composition at first breeding was obtained by breeding gilts at puberty, second, or third estrus and by providing those gilts bred after puberty one of four nutritional regimens from puberty until breeding: 1) 2.7 kg/d of a 14.3% CP, 3.5 Mcal ME/kg diet (H), 2) maintenance ME and CP/d (M), 3) half-maintenance ME and CP/d (1/2M), and 4) M or 1/2M until anestrus, then 2.27 kg/d of a 14.0% CP corn-soybean meal diet until first breeding. Body composition at first breeding was determined using live weight, backfat thickness, and deuterium oxide space as variables in prediction equations. All females were treated similarly after first breeding. Age and body composition at first breeding were not related (P > .10) to litter size at birth or weaning in parities 1, 2, 3, or overall. Increasing age at first breeding was related to small increases in pig birth weights (P < .001) in parity 1 and average pig weaning weight (P < .001) in parities 1, 2, and overall. Body composition of gilts at first breeding was not related (P > .10) to pig birth weights and was inconsistently related to pig weaning weights in parities 2 and 3 (P < .001). Females heavier at first breeding remained heavier (P < .01) throughout the experiment. Age and body composition at first breeding were not different (P > .10) for gilts completing three parities (n = 53) compared with gilts failing to complete three parities (n = 34). Results show no large effects of gilt age or body composition at first breeding on sow productivity and longevity over three parities.