Two relatively fecund genotypes (Flemish Milksheep and KU-Lovenaar-an F 2 cross-bred originating from pure-bred Suffolk and Flemish Milksheep) and two meat-type breeds (Texel and Suffolk) were compared regarding their plasma inhibin and FSH levels. Oestrus was synchronised by progestagen-impregnated sponges. Blood sampling, starting 12 h after the sponge removal, continued twice daily for 21 consecutive days. To localise the pre-ovulatory LH peak and to determine the cycle length, sampling at 2 h intervals was carried out during the first 3 days and the last 4 days of the 21-day period. Immunoreactive inhibin was measured by a 5-24 α-subunit inhibin assay as well as by an assay based on an antiserum against native bovine 31-kDa inhibin (Monash assay). Changes in plasma concentration of immunoreactive FSH and LH during the oestrous cycle agree with the literature data. The profiles generated by the two different inhibin RIAs were different and showed a negative correlation during the periovulatory phase and on days 9-11. No characteristic pattern could be observed during the oestrous cycle. Significant breed differences were observed for 5-24 α-subunit inhibin as well as for inhibin, measured by the Monash assay. The significantly higher (P<0.05) 5-24 α-subunit inhibin and lower inhibin (Monash assay) concentration in Texel ewes resulted in a differential ratio of both inhibin measurements for the Texel in comparison with the other breeds. No significant breed differences for FSH were observed during the oestrous cycle. It is concluded that, although significant breed differences in inhibin concentration are observed, the FSH-inhibin interaction is not the only factor responsible for breed differences in ovulation rate.