Repeated measures of ovulation rate on puberal heifers should be an effective way to select for increased twinning rate. A reliable estimate of the genetic correlation between ovulation and twinning rates is needed to implement such selection and to predict its efficiency. Restricted maximum likelihood was used to estimate genetic correlations from subsets of data collected from the twinning project at the Roman L. Hruska U.S. Meat Animal Research Center. The animal model included numerator relationships among the animals and fixed effects of year-season of measurement, age at measurement, and birth group. Genetic correlations between averages of ovulation rates for three, four, five, six, seven, and eight estrous cycles and the occurrence of twins were, respectively, .62, .76, 1.00, 1.00, 1.00, and .98 based on observations on 200 to 325 puberal heifers. Corresponding phenotypic correlations ranged from .06 to .26. Genetic correlations between ovulation rate in a single estrous cycle and occurrence of twins were .38, .98, and .98, respectively, for 323, 430, and 283 cows with ovulation rate measured after measurement of twinning, in the same season as measurement of twinning, and more than 1 yr before measurement of twinning. Phenotypic correlations were .00, .07, and .00. Genetic and phenotypic correlations among ovulation rates at six consecutive estrous cycles averaged, respectively, .66 and .12 for 610 heifers. Heritabilities for ovulation rates in individual cycles averaged .16. No evidence of negative environmental covariance between ovulation rates in adjacent cycles was found. These results support the approach of indirectly selecting for twinning rate by measuring ovulation rates in estrous cycles of puberal heifers.