The objective of the present study was to determine the optimal culling strategy in relation to biological and economic efficiency (BE and EE, respectively) and annualized net revenue (AN) in the Japanese Black cow–calf production system with special reference to the beef quality of culled cows. The herd model focused on two ways of mating: one-mating trial system (ONE) and continuous-mating trial system (CON). ONE assumed that heifers that fail to conceive are culled and cows that fail to conceive are culled at weaning of their calves, while CON assumed that mating continues until all females theoretically conceive. Least square means of carcass data of Japanese Black cows collected from a cooperative farm in Japan were used to estimate the carcass price of a cow by parity and Beef Marbling Standard (BMS) number. The simulation, assuming the current production situation in Japan, indicated that sales of culled cows accounted for 0·10–0·20 of total sales and was an important element in total production. Comparisons between ONE and CON showed that production efficiency in the current situation is higher in CON. The BE, EE and AN were higher in CON than in ONE. The two economic indicators were less sensitive to changes in annual discount rate but highly sensitive to changes in female calf price and BMS number of cows, indicating the importance of considering fluctuations in calf price and potential quality of culled cows' carcasses when estimating the economically optimal parity of culling. The three indicators derived different optimal solutions even in the same mating trial systems, stressing the importance of choice of production indicators when determining the culling strategy and evaluating animal production.