Reasons for patient non-entry in randomized clinical trials (RCTs) assessing oncologic treatment modalities are not well documented in the literature. We have prospectively recorded reasons for RCT non-entry in breast cancer patients at St. Michael's Hospital, Toronto. From September 1984 to November 1989, 592 consecutive patients were evaluated through the clinical trials office. One hundred six out of the 592 patients were placed into a RCT (17.9%). Protocol ineligibilities accounted for 273 non-entries (46.1%) and protocol eligible but not entered patients accounted for 213 (36.0%) non-entries. The most common reason for protocol ineligibility was advanced age (94 patients). The most common reason for protocol eligible but not entered patients was patient refusal (148 patients). A total of 272 patients in both non-entered groups were identified as having reasons for non-entry that were potentially correctable. In summary, protocol ineligibilities account for the majority of non-entered patients, but patient refusal accounted for the single largest group of potentially correctable non-entries. More dissemination about the merit of RCTs in the lay press and amongst primary care physicians must take place if we are to expediently and efficiently answer important oncologic questions.