Network theory conceptualizes the organization of semantic information as nodes within a net-like structure. Links within the network represent associations among knowledge structures or concepts and have been viewed as forming the basis of human semantic memory (Quillian, 1962). Network models have recently been employed in emotion theory (Bower, 1981; Lang, 1984) and in research in human sexuality (Geer, 1996; Rabalais & Geer, 1996; Smith, Eggleston, Gerrard, & Gibbons, 1996). As a result of that research, stable gender differences in the organization of knowledge for sexual and emotional information have begun to be identified (see Geer & Manguno-Mire, 1996, for a review). These differences in knowledge organization have been shown to map well onto existing research demonstrating gender differences in sexual attitudes, fantasies, and behavior (Geer & Manguno-Mire). The present study employs the Pathfinder computer algorithm (Schvaneveldt, 1990) to compare the cognitive associative networks of heterosexual men and women and gay men and lesbians. Ninety homosexuals (49 gay men and 41 lesbians) and 95 heterosexuals (48 men and 47 women)rated all pair-wise combinations of 16 words relevant to sexuality and emotion. The sample was predominantly Caucasian (90%). Four percent of participants were African-American, 3% were Hispanic American, 1% were Asian American. Analysis of our dependent variable (number of links on relevant concepts) revealed gender and sexual orientation differences in the total number of links in associative networks, the number of links within relevant sexual and emotional word clusters, between relevant word clusters, and on individual words. Results are interpreted in light of relevant theories of gender and sexual orientation.