None of the commercial HPV tests are U.S. FDA-approved for testing of cervical cytology specimens in SurePath preservative. Still, ~30% of HPV testing is performed on specimens in this formalin-containing preservative. Formalin-induced DNA fragmentation and cross-linking may interfere with HPV detection. We evaluated analytical sensitivity and specimen stability of the cobas 4800 HPV (Roche) and Hybrid Capture 2 HPV (HC2, Qiagen) tests with residual cervical cytology samples in SurePath preservative available within 1 week of collection. Cobas testing was performed with and without heating samples at 120°C for 20 min diluted 1:1 in an alkaline environment (pretreatment) to revert DNA crosslinking. Stability was tested after 2 weeks of storage at ambient temperature followed by ≤10 weeks at 4°C. Analytical sensitivity and positivity rates (HC2, 18%; cobas pretreated, 46%; cobas untreated, 47%) were greater for cobas than HC2 (n = 682). After 6 weeks of storage, mean HC2 ratios were lower (mean 0.9, SD 6.3) but high variability limited statistical power to detect trends. Cobas threshold cycles (Ct's) increased in untreated (mean 2.1) but not pretreated samples (mean 0.3; n = 110; p≤0.0001). Overall, cobas had greater analytical sensitivity for samples in SurePath preservative. Although pretreatment introduced a manual sample transfer step and 30 min of incubation times, it improved stability without negatively affecting analytical sensitivity. While awaiting results of large trials to evaluate the clinical performance of cobas, the addition of the pretreatment step may improve the detection of HPV, especially after prolonged sample storage.