Consistent with our hypotheses concerning partnership type, bisexually-identified men were
less likely to report being involved with main male partners exclusively (19% bisexually identified, 36% gay identified); they were also more likely to report being involved with only nonmain male partners (54% bisexuallyidentified, 45% gay-identified) or being involved with both main and nonmain male partners (26% bisexually-identified, 19% gay-identified).