Abstract Background Little comparative data on substance use (SU) between sexual minority youth (SMY) and heterosexual youth (HET) is available. This study compares the prevalence of SU in an urban cohort between SMY and HET and evaluates demographic and psychosocial predictors of SU. Methods Data came from a prospective-longitudinal cohort study in an urban setting (N = 1297). SU and psychosocial variables such as internalizing symptoms, self-control, sensation-seeking, bullying-victimization, subjective stress, leisure activities, and peer influences were assessed with self-reports at age 17 and 20. SU was stratified by sex and sexual attraction, and the groups were compared using regression models, with demographic and psychosocial variables included as covariates. Results SMY- and HET-youth displayed differences in a number of psychosocial variables. Overall, SMY- and HET-youth differed in their 12-months prevalence of SU: At age 17, SMY-females had significantly higher rates of SU than HET-females for cannabis (aOR = 2.14, p = 0.04), ecstasy/MDMA (aOR = 4.29, p = 0.01), and hallucinogens (aOR = 5.59, p = 0.02). At age 20, SMY-females had significantly higher rates of SU than HET-females for tobacco (aOR = 2.06, p = 0.03), cannabis (aOR = 2.24, p = 0.004), ecstasy/MDMA (aOR = 3.93, p < 0.001), stimulants (aOR = 3.45, p = 0.002), and hallucinogens (aOR = 6.65, p < 0.001). SMY-males reported significantly lower rates for tobacco and cannabis than HET-males at age 17. At age 20, they reported significantly higher rates for the use of ecstasy/MDMA (aOR = 2.30, p = 0.04) and hallucinogens (aOR = 2.43, p = 0.03). Conclusions Given that psychosocial variables were significant covariates of SMY-status and SU, our results underline the importance of accounting for these when explaining differences in SU between adolescents. While differentiation by sex is established in most studies, such standardized comparisons are lacking with regards to sexual identities. But knowledge about SU of SMY is critical for designing effective interventions. This is especially true for SMY-females: Thus, SU in SMY-females early in life needs to be explored more thoroughly and addressed with adequate prevention measures.