Objectives: To assess the determinants of change in clinical practice patterns of physicians with respect to the use of ovulation inductio drugs and the potential for an increased risk of ovarian cancer.Materials and Methods: A questionnaire was mailed to 411 randomly chosen obstetricians and gynaecologists in English-speaking Canada in mid-1996. The response rate was 72%. The main outcome measures were demographics, extent of involvement in fertility treatment, knowledge of the medical literature; information sources, and practice patterns.Results: All but 26 respondents practiced at least some infertility therapy, but only a small proportion (21%) devoted more than 25% of their practice to infertility treatment. Ninety-six percent of respondents had heard of the controversy about fertility drugs and ovarian cancer and of these, 23 felt the link was either non-existent or very weak. Forty percent of physicians routinely discussed the topic of ovarian cancer with their patients before prescribing fertility drugs and for 70% of them, this was something they had only begun to do routinely in the year previous to the survey. Twenty-six percent of respondents had altered their prescription pattern with respect to the maximum total number of treatment cycles because of the potential risk of ovarian cancer. Standard demographics were unrelated to practice change. Those physicians who used gonadotropins were more likely to have changed their practice than were those who used clomiphene alone. Those who had changed their prescription patterns were influenced by the existing evidence, medical conferences and CME events, while pharmaceutical representatives and fear of litigation had little affect. Those who had changed their practice patterns seemed to be more familiar with the literature and methodological issues; 68% realized that the majority of the current evidence was based on cohort studies and case control studies, as compared with 44% of those who hadn't made any changes in their practice. However, they were also more likely to think there was a dose-response relationship between fertility drugs and ovarian cancer and to be satisfied with the number of studies that have been done. Physicians who had changed their prescription patterns were less likely to treat those with a family history of breast or ovarian cancer or those who had already completed at least 12 cycles with fertility drugs, and more likely to treat patients who had previously used oral contraceptives.Conclusions: Virtually everyone had heard about the possible link between fertility drugs and ovarian cancer, but only 40% routinely discussed the controversy with their infertility patients. Those physicians who had lowered their limit for maximum number of treatment cycles believed there was good evidence to support their practice changes, but most appeared not to be aware of methodolgical weaknesses in the research. Future studies should aim to determine how best to disseminate the most current information to physicians.