This study was conducted with random samples of pigs of a line (LS) selected for eight generations for litter size and a randomly selected control line (RS) in the Nebraska Gene Pool population and of crosses of these lines with a Large White-Landrace composite line (I) that was selected for an index of ovulation rate and embryonic survival to 50 d of gestation. Two experiments were done to obtain information on ovulation rate, uterine dimensions, and uterine capacity. In Exp. 1, 103 gilts were slaughtered 9 to 16 d after their second estrus to obtain reproductive tracts for evaluation of ovulation rate and uterine dimensions. In Exp. 2, unilateral hysterectomy-ovariectomy was performed 3 to 10 d after puberty in 109 gilts that were then mated at their next estrus and slaughtered at 93 to 100 d of gestation to recover reproductive tracts for evaluation. Litter size at birth was recorded at first (n = 414), second (n = 159), and third (n = 143) parity of pureline and crossline gilts. The cumulative response to eight generations of selection for litter size pooled over type of cross and parity was 1.21 +/- .38 pigs, in good agreement with the realized response of 1.06 pigs previously estimated for the period of selection. This response was due to an increase of 1.30 +/- .54 eggs in ovulation rate (measured by the number of corpora lutea in cyclic and pregnant gilts) and .66 +/- 1.28 pigs in uterine capacity (measured after unilateral hysterectomy-ovariectomy). No significant changes were found in uterine dimensions in cyclic gilts. Approximately 25% of the increase in litter size could be explained by a reduction in number of mummified pigs at birth, an indication that uterine capacity in late gestation was increased. Estimated differences between lines (gLS - gRS) were not significantly different for pureline and crossline gilts and sows, and no interactions of these estimates with parity were detected. Estimates of heterosis for LS x I and RS x I did not differ significantly, nor were interactions of heterosis with parity significant.