The effects of pollinator density on the seed set of a male-sterile ( MS) and a male-fertile (MF) line of winter oilseed rape ( Brassica napus) were studied in 24 isolation cages ( 7.5 m(2)). Pollinator treatments comprised of high pollinator density ( a small honeybee colony and 10 mason bees, Osmia rufa), and a gradient of increasing O. rufa densities from zero ( control) up to 100 bees per cage. High pollinator densities increased the seed weight per plant from 6.5 g to 56.9 g for the MS line. However seed weight did not differ between high pollinator densities and controls for the MF line. Increasing densities of O. rufa had a significant effect on almost all yield components of the MS line, but only marginal effects on the MF line. The number of seeds per pod, the seed weight per plant and the harvest index ( seed weight/plant dry weight) of the MS line showed a significantly steeper increase with increasing bee density than that of the MF line. The results suggest that solitary bees could be used successfully to replace honeybees as pollinators of MS oilseed rape in isolation cages.