Reproductive efficiency of dairy bulls is usually measured by nonreturn rate. Nonreturn is a compound trait that is a result of two events, conception and gestation, that lead to calving. Nonreturn rates can be used to derive more elementary biological measures for reproductive efficiency, such as conception rate and calving rate, which separately might be more reliable than nonreturn rate itself to evaluate the fertility of a bull or the performance of an AI technician. The challenge of this study was to examine the decline in nonreturn rate in light of the theory of multiphasic analysis. A multiphasic logistic function was developed to model decline in nonreturn rate by estimating conception rate, calving rate, and characteristics of the first two estrous cycles. The model is illustrated with data on daily nonreturn rates to 120 d. From the proportion of cows that conceived but failed to complete gestation because postsignal embryonic death, the model estimates conception rate and calving rate. From the proportion of cows that failed to conceive or that conceived but failed to complete gestation because of presignal embryonic death, the model estimates the proportion of returns, or probability of detecting estrus, duration of nonreturns, and time of maximum decline in nonreturn rate for the first two cycles. Using the proposed model, conception rate and calving rate estimated from daily nonreturn rates might be more reliable for evaluation of performance of an AI technician and fertility of a bull than nonreturn rates at arbitrarily chosen days after insemination.