This study uses the 2000 U.S. Census data to assess the impact of antidiscrimination policies for sexual orientation on earnings for gays and lesbians. Using a multilevel model allows estimation of the effects of state and local policies on earnings and of variation in the effects of sexual orientation across local labor markets. The results suggest that gay men face an earnings penalty that varies significantly (though not sizably) across local areas, and that state antidiscrimination policies may decrease that penalty in private sector employment. There is, however, no evidence that lesbians in any sector average higher earnings or wages in areas with antidiscrimination policies. The strongest evidence of effects for antidiscrimination policies is for weeks of employment and for gay men who are in the private sector, white, and in the upper half of the earnings distribution. (Contains 6 tables, 4 figures and 37 footnotes.)