The flowering racemes of Brassica napus L. cv. Tapidor were exposed independently from the vegetative component to 200 nl l−1, 100 nl l−1 O3/30 nl l−1 SO2, In all cases, a single 6 h exposure significantly increased bud abortion, and abscission 2 and 5 d after treatment. Similar results were obtained using cv. Libravo, although the effect of 100 nl l−1 O3 was significant only after 2 d. Exposure to 30 nl l−1 SO2 did not enhance bud abortion and abscission in either variety. Longer term measurements for up to 25 d revealed a trend to wards sustained losses of fertile sites in those pollutant treatments which had exhibited demonstrable, effects after 2 and/or 5 d. However, the number of fertile sites present was comparable to or even higher than in the controls, indicating that compensation for the initial losses must have occurred. Following exposure, plants of cv. Tapidor tended to develop longer total raceme lengths and greater numbers of raceme branches, and those of cv. Libravo a greater number of raceme branches. Indicating possible methods by which new fertile sites may have been produced. The significance of these observations for the reproductive development of field-grown oilseed rape is discussed.