AbstractA field experiment was conducted to assess the effect of autumn defoliation treatments on inflorescence production, potential seed yield and yield components of white clover cv Makibashiro. Between 10 July and 10 October 1992, white clover swards were subjected to one of three treatments: monthly cutting to 3–4 cm (4DEF), cutting to 3–4 cm on 10 August and 10 October (2DEF), and no cutting (control). The total numbers of inflorescences and the proportion of inflorescences in different development categories were counted throughout the period of seed crop development to determine the pattern of inflorescence development and optimum harvest date. There were consistent significant differences in inflorescence density between treatments. Plots which received the 2DEF treatment produced significantly more inflorescences than did the no‐defoliation (control) and 4DEF‐treatment plots. In this particular year the optimum harvest date (date at which the proportion of ripe inflorescences and potential seed yield was highest) was 26 July, approximately 30 d after peak flowering. Defoliation treatments had no effect on optimum harvest date. However, treatments differed in potential seed yield and ripe inflorescences on this date. The 2DEF treatment gave the highest potential seed yield because there were significantly more ripe inflorescences than either the contol or 4DEF plots. Control plots produced inflorescences with more florets than the other two defoliation treatments, but the differences were not always significant. Seed number per pod was higher in inflorescences obtained from previously defoliated plots than from control plots. The 1000‐seed weight was significantly lower in inflorescences developed in 4DEF plots than those developed in 2DEF and control plots. The results are discussed in relation to the management of white clover seed crops and the importance of canopy structure and light intensity for seed production.