Two modifications to the original L1 consensus primer human papillomavirus (HPV) PCR method, MY09–MY011, using AmpliTaq DNA polymerase (MY-Taq), were evaluated for HPV DNA detection on clinical specimens from a cohort study of cervical cancer in Costa Rica. First, HPV DNA testing of 2978 clinical specimens by MY09–MY011 primer set, using AmpliTaq Gold DNA polymerase (MY-Gold) were compared with MY-Taq testing. There was 86.8% total agreement (κ = 0.72, 95%CI = 0.70–75) and 69.6% agreement among positives between MY-Gold and MY-Taq. MY-Gold detected 38% more HPV infections (P < 0.0001) and 45% more cancer-associated (high-risk) HPV types (P < 0.0001) than MY-Taq, including 12 of the 13 high-risk HPV types. Analyses of discordant results using cytologic diagnoses and detection of HPV DNA by the Hybrid Capture 2 Test suggested that MY-Gold preferentially detected DNA positive specimens with lower HPV viral loads compared with MY-Taq. In a separate analysis, PGMY09–PGMY11 (PGMY-Gold), a redesigned MY09/11 primer set, was compared with MY-Gold for HPV DNA detection (n = 439). There was very good agreement between the two methods (κ = 0.83; 95%CI = 0.77–0.88) and surprisingly no significant differences in HPV detection (P = 0.41). In conclusion, we found MY-Gold to be a more sensitive assay for the detection of HPV DNA than MY-Taq. Our data also suggest that studies reporting HPV DNA detection by PCR need to report the type of polymerase used, as well as other assay specifics, and underscore the need for worldwide standards of testing. J. Med. Virol. 68:417–423, 2002. © 2002 Wiley-Liss, Inc.