A PROGRAMME of breast self-examination (BSE) education was organised for women aged 45 to 64 years of age in a health district. Women who attended completed a pre- and post-teaching ques tionnaire, and were given a calendar card on which to record their BSE practice in the year following education. The calendar was recalled and served as a measure of the degree of change in BSE frequency from that claimed on the questionnaire. The rate of return was lower than expected. Compared to wom en who did not return a calendar, those who did were more aware of the programme, BSE, and causes of breast lumps; and more of them claimed to have practised BSE before education. A high proportion of women who returned calendars maintained or achieved satisfactory frequency of BSE practice in the year following education. For this group of women the calendar was a satisfactory measure of behaviour change.