In a closed population and defined time period, the mean number of opposite-sex partners reported by men and women should be equal. However, in all surveys, men report more partners. This inconsistency is pivotal to debate about the reliability of self-reported sexual behavior. We used data from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3), a probability sample survey of the British population, to investigate the extent to which survey sampling, accounting strategies (e.g., estimating versus counting), and (mis)reporting due to social norms might explain the inconsistency. Men reported a mean of 14.14 lifetime partners; women reported 7.12. The gender gap of 7.02 reduced to 5.47 after capping the lifetime partner number at the 99th percentile. In addition, adjusting for counting versus estimation reduced the gender gap to 3.24, and further adjusting for sexual attitudes narrowed it to 2.63. Together, these may account for almost two-thirds of the gender disparity. Sampling explanations (e.g., non-U.K.-resident partners included in counts; sex workers underrepresented) had modest effects. The findings underscore the need for survey methods that facilitate candid reporting and suggest that approaches to encourage counting rather than estimating may be helpful. This study is novel in interrogating a range of potential explanations within the same nationally representative data set.