SUMMARY By means of discriminant analysis the validity of a fourfold qualitative classification of vaginal smears has been examined in two inbred lines of mice and their reciprocal crosses. For both the intravaginal and subcutaneous routes of administration with oestradiol-3:17β the vaginal response is shown to be truly quantal, no graded classification of responses being necessary. It is shown that prooestrous smears should be classified as positive and also that the presence or absence of leucocytes is the best criterion for classification.