Background The role of circumcision in male HPV acquisition is not clear. Methods Male university students (18–20 years of age) were recruited from 2003–2009 and followed tri-annually. Shaft/scrotum, glans, and urine samples were tested for 37 alpha HPV genotypes. Cox proportional hazards methods were used to evaluate the association between circumcision and HPV acquisition. Logistic regression was used to assess whether number of genital sites infected at incident HPV detection or site of incident detection varied by circumcision status. Results In 477 men, rates of acquiring clinically-relevant HPV types (high-risk types plus types 6 and 11) did not differ significantly by circumcision status (hazard ratio [HR] for uncircumcised relative to circumcised subjects: 0.9[95%CI:0.7–1.2]). However, compared to circumcised men, uncircumcised men were 10.1 (95%CI:2.9–35.6) times more likely to have the same HPV type detected in all 3 genital specimens than in a single genital specimen and were 2.7 (95%CI:1.6–4.5) times more likely to have an HPV-positive urine or glans specimen at first detection. Conclusions While the likelihood of HPV acquisition did not differ by circumcision status, uncircumcised men were more likely than circumcised men to have infections detected at multiple genital sites, which may have implications for HPV transmission.