We measured aspects of community involvement chosen for men who have sex with men (MSM) in Sweden (gay places, media accessed, Internet, gay festivals, and social engagement, measured as proportion of gay friends) in two Swedish Internet-based samples from 2006 (n=3,202) and 2008 (n=4,715). Data showed low to moderate reliability with a moderate (0.57) alpha coefficient. While there is moderate internal consistency, as might be anticipated from measures of actual community involvement, they can be treated as scales. The Internet scale indicated the lowest reliability, perhaps due to respondents having Internet sites of primary choice, rather than a high level of usage across several sites. A hypothesized lack of correlation between traditional domains of the gay community and the Internet did not appear: correlations between the Internet measure and the other measures were positive and significant, but among the lowest correlations obtained between the community measures, ranging from 0.06 to 0.24. Those who use the Internet extensively are less likely to be involved in other aspects of the community. Sexual risk was associated with high social engagement at sexual meeting sites and with Internet use. Gay community involvement, including the Internet community, may be complex and associated with both increase in HIV sexual risk behaviors (by measuring use of sexual risk sites) and preventive measures (HIV testing).