Researchers have suggested the rising use of Internet to look for sexual partners is an important contributor to the resurgence in the incidence of sexually transmitted diseases (STDs) and HIV among men who have sex with men (MSM). Web-based samples of this population reflect good geographical diversity for research, but a high drop-out rate creates a significant potential for bias, misrepresentation of population, and misinterpretation of data. This study aims to describe the demographics of an Internet sample of MSM and the differences between the completers and drop-outs. We analyzed data obtained from a cross-sectional study using an online self-administered questionnaire for males, 18 years or older, who were current U.S. residents, and who had sex with men. Of 850 eligible participants, 404 (47.5%) were labeled as drop-outs. The completers and drop-outs differed significantly in age, education, country of birth, health insurance, time spent on Internet, location of computer access, types of sites visited, profile information, last homosexual experience, methods used to determine safety of partners, and type of sexual activities on real meets. Almost half of the participants dropped out before completion. Although the two groups did not differ in many of the items, high drop-out rates are a threat to the validity of such data. Drop-out constitutes a significant bias in Internet sexuality research and must be considered while interpreting the results of such studies.