The present study was conducted to investigate whether white clover (Trifolium repens L.) in grazed swards in Uruguay can produce and shed sufficient seed, and whether sufficient seedlings can develop and survive, to make an important contribution to the long-term persistence of the species. Five sites, on which survival of white clover had been better than is usual in Uruguay, were studied in 1995–97. The sites had been sown with white clover 11–15 years previously and were managed mainly by cattle grazing. The number of white clover inflorescences/m2 was at least 18 at nearly half the counts in the study and the number of viable seeds in the top 5 cm of soil was at least 1500 at half the counts. The results for both the above variables suggest that all five sites had a supply of seed greater than a commercial seed rate at some times of the year. However, the number of seedlings/m2 was >100, a possible target in these conditions, at only one fifth of the counts, and the proportion of seedlings that survived from one recording date to the next (a period of 4–5 weeks) varied from >0·7 to zero. It seems, therefore, that, in these conditions, the failure of seeds to convert to established plants is likely to be a greater constraint than the production and shedding of seeds. There was considerable variability with time, and between and within sites at each stage of the reproductive cycle. On one site, competition from Bermudagrass (Cynodon dactylon) greatly reduced the number of white clover seedlings/m2. Overall, the results suggest that white clover in grazed swards in Uruguay can produce and shed sufficient seed, and that sufficient seedlings can develop and survive, to make a significant contribution to the long-term persistence of the species. However, each year there is a risk of seedlings being killed in dry, hot weather.