Recent studies have suggested that long-term oxytocin administration can alleviate the symptoms of autism spectrum disorder (ASD); however, factors influencing its efficacy are still unclear. We conducted a single-center phase 2, pilot, randomized, double-blind, placebo-controlled, parallel-group, clinical trial in young adults with high-functioning ASD, to determine whether oxytocin dosage and genetic background of the oxytocin receptor affects oxytocin efficacy. This trial consisted of double-blind (12 weeks), open-label (12 weeks) and follow-up phases (8 weeks). To examine dose dependency, 60 participants were randomly assigned to high-dose (32 IU per day) or low-dose intranasal oxytocin (16 IU per day), or placebo groups during the double-blind phase. Next, we measured single-nucleotide polymorphisms (SNPs) in the oxytocin receptor gene (OXTR). In the intention-to-treat population, no outcomes were improved after oxytocin administration. However, in male participants, Clinical Global Impression-Improvement (CGI-I) scores in the high-dose group, but not the low-dose group, were significantly higher than in the placebo group. Furthermore, we examined whether oxytocin efficacy, reflected in the CGI-I scores, is influenced by estimated daily dosage and OXTR polymorphisms in male participants. We found that >21 IU per day oxytocin was more effective than ⩽21 IU per day, and that a SNP in OXTR (rs6791619) predicted CGI-I scores for ⩽21 IU per day oxytocin treatment. No severe adverse events occurred. These results suggest that efficacy of long-term oxytocin administration in young men with high-functioning ASD depends on the oxytocin dosage and genetic background of the oxytocin receptor, which contributes to the effectiveness of oxytocin treatment of ASD.