This study sought to identify factors that facilitate or hinder participation in a breast cancer health promotion trial among high-risk women. The subjects were 271 women ages 35 years and older who had a family history of breast cancer in at least one first-degree relative. All subjects were eligible for participation in a randomized trial which compares breast cancer risk counseling with general health counseling. Structured telephone interviews evaluated demographic characteristics, risk factors, risk perceptions, breast cancer concerns, and past screening practices. The results showed that education level was a key determinant of the importance of these factors in participation. Logistic regression modeling indicated that women with a high school education or less were most likely to participate if: (a) their relatives' diagnoses had greatly increased their perceptions of their personal risks [OR (OR) = 4.1], particularly if they perceived that risk to be very high (OR for interaction = 6.4); and (b) if they were ages 40-49 years versus 35-39 or 50 + years (OR = 2.6). By contrast, among women with education beyond high school, participation was predicted by (a) marital status (OR = 2.6), (b) employment (OR = 0.03 for employed), (c) number of affected relatives (OR = 0.07 for 1 versus 2 first-degree relatives), and (d) previous biopsy (OR = 0.42). These findings suggest that recruitment strategies that tailor messages to women's educational levels might be most effective.