In the past decades, we have seen an increased interest in including sexual and gender minorities in surveys, which allow for the estimation of the size of these groups and which allow inference to larger populations of persons with LGBT identities. This is incredibly tricky, if not impossible, as no exhaustive list of persons with LGBT identities exist, which could serve as a sampling frame. In addition, there are other challenges associated with collecting quality survey data on these populations, such as the small group size relative to the heterosexual cis-majority. I will use two recent survey projects, in which colleagues and I have used different strategies to collect quality data among LGBT households in Germany (SOEP-LGB project 2019) and persons in same-sex couples the Netherlands (UNICON project 2016), to illustrate some of the possible strategies and their trade-offs. In a second part, I discuss two empirical studies based on the Dutch UNICON data in more detail. The first study tests the hypothesis of weaker intergenerational ties between parents and their adult lesbian daughters and gay sons compared to heterosexuals. Thereby, we directly test mechanisms linking the strength of these ties to gender differences and the liberal or traditional views held by the parents when the child was growing up. We also include the process of leaving home in the analyses. The second study, examines sexual orientation as stratifying factor in social networks. Specifically, it is a comparison of the size, composition and density of social networks.