Incidences of pollination and fertilization were determined and the development of seeds was monitored in spring beans over 2 years. Plant density was varied in the 1st year and irrigation in the 2nd. The incidences of pollination, fertilization and seed development were all independent of plant density. The proportion of ovules that developed as seeds was slightly enhanced in irrigated plots as losses from abortion and from lack of fertilization within fertilized ovaries were both reduced. Autofertility, or spontaneous self-pollination leading to fertilization, led to an increase in the overall incidence of fertilization. Three autofertile stocks were up to 11% more pollinated and fertilized than two partly autofertile ones which, in turn, were up to 20% more than two non-autofertile, long-podded stocks. A closed-flower population was 30% less fertilized than the long pods and was insufficiently autofertile to allow an adequate number of pods to set on the lower flowering nodes. Three stocks, one autofertile, one with terminal inflorescences (ti) and the third with both characters, were subject to particularly high losses of fertilized ovules from abortion, 27–35%, while few ovules, 12% or less, aborted in closed flowers and long pods. Failure of seeds to mature in long pods was associated primarily with the low proportion of ovules fertilized within fertilized flowers, rather than with abortion of the fertilized ovules.