A project was implemented in 1981 with the objective of increasing twinning rate in cattle. Daughters of foundation sires had twin calves at a frequency of from 8 to 13%, and foundation females had twin calves at an average frequency of 50%. Data were analyzed on twinning rate, ovulation rate, and weight traits. The h2 of ovulation rate increased from .11 to .38 for a single estrous cycle to the mean of eight estrous cycles. From all data, h2 for single observation of ovulation rate and twinning rate were .10 and .09, respectively. The r(g) between them was .75. The h2 of weight traits ranged from .42 to .54 when weight traits were analyzed pair-wise with ovulation rate and with twinning rate. The r(g) between weight traits with ovulation rate ranged from .15 to .30 and with twinning rate ranged from .24 to .39. Phenotypic mean twinning rate increased from 1.07 to 1.29 calves per parturition for females born 1981 through 1993, and adjusted mean predicted breeding value (PBV) increased from 1.07 to 1.33 calves per parturition. Phenotypic mean ovulation rate for fall of 1984 through fall 1994 birth groups increased from 1.11 to 1.26 per estrous cycle, and adjusted mean PBV for ovulation rate increased from 1.11 to 1.29 per estrous cycle. Because of the high r(g) (i.e., .75) between ovulation rate and twinning rate, and because of a h2 of .35 for ovulation rate for the mean of six estrous cycles, repeated records of ovulation rate in puberal heifers is an effective indirect selection criterion for twinning rate. The positive r(g) between growth traits and ovulation and twinning rate suggest the need for some compromise when the selection goal is increased twinning rate with no increase in growth and size.