A stochastic computer simulation model was used to predict heritability (h2) and repeatability (t) of derived reproductive traits in beef cattle as a function of underlying, normally-distributed genetic variation in interval from calving to first estrus (postpartum interval) and single-service conception rate. Traits simulated were mating rate, first-service conception rate, overall conception rate, date of first service, number of services during a 63-d breeding season, calving date and postpartum interval. Weight of calf weaned was also simulated on a per-cow-exposed, per-cow-mating, per-cow-calving, per-calf-weaned or adjusted 205-d basis. Derived estimates of h2 for mating rate and date of first service were not significantly different from 0, reflecting the high proportion of cows that were predicted to cycle within the first 21 d of breeding. Estimates of h2 were significant and approached the level of input h2 for overall and first-service conception rate and number of services, suggesting that these traits may be potential selection criteria. Values of h2 for calving date were significant but considerably smaller than input h2. Estimates of t were much larger than h2 for date of first service and calving date, even when no nonadditive genetic or permanent environmental effects were explicitly simulated. Estimates of h2 for weaning weight per cow exposed and especially for weaning weight per cow calving (in the presence of random calf death losses) were much lower than underlying h2 values. Culling of open cows generally reduced genetic variances of derived reproductive traits to negligible levels within three calf crops.