Purpose: The purpose of this study was to assess the feasibility and efficacy of using advertisements (ads) on Facebook, Instagram, and Snapchat to recruit a national sample of adolescent sexual minority males ages 13-18 and transgender youth ages 13-24 for a web survey. Methods: The Survey of Today's Adolescent Relationships and Transitions (START) used targeted ads as survey recruitment tools. We assessed the efficacy of these varied forms of recruitment ads in reaching our target population. To understand how our sample differed from a national probability sample targeting the general adolescent population, we compared START respondents with sexual minority men identified from the 2017 Youth Risk Behavior Survey (YRBS). Results: The use of targeted language produced higher rates of completes per click compared with ads without targeted language. Video ads (compared with static images) were more effective at recruiting younger respondents. START and YRBS samples differed along lines of sexual identity, race and ethnicity, and age. The START sample had a greater percentage of Hispanic/Latino and Other/Multiracial respondents relative to the YRBS sample, thus providing additional data on these underserved sexual minority youth. Conclusion: The factors associated with design decisions for a hard-to-reach, non-probability sample impact the likelihood that respondents engage in and complete a survey. The ads proved to be effective and efficient at recruiting the targeted population.