Recent studies have suggested a relationship between bovine follicular fluid testosterone concentration and the likelihood of the oocyte being fertilised by an X- or Y-bearing spermatozoon; however, this theory has been challenged. To further test this hypothesis, follicles were dissected from the ovaries of slaughtered heifers, measured and carefully ruptured. The cumulus–oocyte complex (COC) was removed and the follicular fluid collected and testosterone concentration determined by radioimmunoassay. COCs were matured, fertilised and cultured in an individually identifiable manner; all cleaved embryos (2- to 4-cell stage, n = 164) had their sex determined by PCR. Testosterone concentrations were positively skewed. There was no significant difference between follicular fluid testosterone concentrations in male and female embryos (mean ± s.e.m. 51.5 ± 5.59 and 49.5 ± 7.42 ng mL–1, respectively). Linear, quadratic and cubic logistical regression showed that follicular testosterone concentration could not reliably predict the sex of the embryo with odds ratios of 1.001, 1.013 and 1.066, repectively, and coefficient of determination (R2) values of 0.0003, 0.0126 and 0.0567, respectively. Follicular size and testosterone concentration were not related (R2 = 0.087). Finally, follicular size had no influence on embryo sex determination (P = 0.70). In conclusion, under the conditions of the present study, the likelihood of an oocyte being fertilised by an X- or Y-bearing spermatozoon was not affected by the size of the follicle from which it was derived, nor by the testosterone concentration in the follicular fluid.