Abstract Despite its increasing use, few studies have reported on demographic representativeness and costs of research recruitment via social media. It was hypothesized that cost, reach, enrollment, and demographic representativeness would differ by social media recruitment approach. Participants were 18–25 year-olds at moderate to high risk of skin cancer based on phenotypic and behavioral characteristics. Paid Instagram, Facebook, and Twitter ads, unpaid social media posts by study staff, and unpaid referrals were used to recruit participants. Demographic and other characteristics of the sample were compared with the 2015 National Health Interview Survey (NHIS) sample. Analyses demonstrated significant differences among recruitment approaches regarding cost efficiency, study participation, and representativeness. Costs were compared across 4,274 individuals who completed eligibility screeners over a 7-month period from: Instagram, 44.6% (of the sample) = 1,907, $9 (per individual screened); Facebook, 31.5% = 1,345, $8; Twitter, 1% = 42, $178; unpaid posts by study staff, 10.6% and referred, 6.5%, $1. The lowest rates of study enrollment among individuals screened was for Twitter. Most demographic and skin cancer risk factors of study participants differed from those of the 2015 NHIS sample and across social media recruitment approaches. Considering recruitment costs and number of participants enrolled, Facebook and Instagram appeared to be the most useful approaches for recruiting 18–25 year-olds. Findings suggest that project budget, target population and representativeness, and participation goals should inform selection and/or combination of existing and emerging online recruitment approaches.