By the end of 1972, a total of thirteen rabbits had been injected. One serum was raised to Brassica oleracea S23S23 kale pollen proteins, but no s-antibody was stimulated. All the others were raised to stigma proteins. A range of s-alleles and varieties of B.oleracea was used including S23S23, S16S16 kale, S2S2, S45S45 and S5S5 brussels sprout and S15S15 cabbage. S15 and S5 were of low dominance, S2 and S45 of intermediate dominance and S23, S14 and S16 of high dominance. S-antibodies were raised to S23 and S16 in kale, and S2 and S45 in brussels sprout. Only the S16-antibody had a high titre of 1/16, the other were 1/4 or less. In two of the cases , booster injections stimulated the S-antibody response where the first course of injections stimulated the S-antibody response where the first course of injections stimulated the S-antibody response where the first course of injections had failed. Relatively dilute stigma extracts of 250 stigmas/ml stimulated most s-antibodies. An injection schedule of more than two months produced cross-reactions or new specificities. Extract which had been frozen or treated with formalin gave a poorer antibody response than freshly-prepared extract. No S-antibodies were stimulated by these treatments. In its present form, the technique was unsuitable for routine S-allele diagnosis in B.oleracea because of the low rate of successful sera and the low titres stimulated. Improvements to the technique are suggested. Between 35 and 40% of the total protein content was lost when a stigma extract was frozen, but the S-protein was not lost. B.oleracea pollen germinated on an agar and sucrose medium. Germination was stimulated by 0.0005% queroetin, but tube growth was not. Querotin was detected as glycosides in the pollen and stigma tissue of B.oleracea. It was not involved in the incompatibility reaction of the stigma.