Previous work, in which female mice had been given fractionated doses of 20 X 10 rad X-rays, had confirmed and extended Russell's observations that the dose-response relationship for specific-locus mutations in mature-mouse oocytes is curved at low doses. The present work was intended to study the relationship at relatively high doses. Adult female mice were given doses of 200, 400 or 600 rad x-rays at 52 or 72 rad/min, and mated immediately. Offspring conceived in the first 7 days (i.e. using oocytes which were mature at time of treatment) were scored for specific-locus mutations. The data indicate that the departure from linearity of the dose-response curve is marginally significant at the 5% level. A quadratic dose-response curve (y = c + aD + bD2) and a square-law relationship (y = c + bD2) both give a good fit to the data. Both curves fit data of other authors obtained at low doses or dose-rates. These results could be interpreted either in terms of dose-dependent repair phenomena, or by considering specific-locus mutations as two-track events. In view of knowledge of other phenomena concerning mutation and cell killing in mouse oocytes, such as the variation in sensitivity of different cell stages, the interpretation in terms of repair phenomena is preferred.