We determined the feasibility of human papillomavirus (HPV) detection in cervical exfoliated cells collected as dry swab samples. Both dry cervical swab and specimen transport medium (STM) cervical swab samples were collected from 135 patients attending either colposcopy or women's clinics in Guayaquil, Ecuador, who had a cytology diagnosis within 6 months. HPV was detected by dot blot hybridization and genotyped by the liquid bead microarray assay (LBMA). Overall, 23.1% of dry samples were positive for any high-risk HPV types, and 24.6% of STM samples were positive for any high-risk HPV types. Of 125 paired samples, the type-specific high-risk HPV proportion positive agreement was 60.7% (kappa, 0.69; 95% confidence interval [CI], 0.53 to 0.82). Of six women with cytological evidence of invasive cervical cancer, high-risk HPV DNA was detected in three of their STM samples and in five of their dry samples. Dry samples were more likely to be insufficient for HPV testing than STM samples. Consistent with this observation, the amount of genomic DNA quantitated with the beta-actin gene was almost 20 times lower in dry samples than in STM samples when detected by the real-time TaqMan assay; however, HPV DNA viral loads in dry samples were only 1.6 times lower than those in matched STM samples. We concluded that exfoliated cervical cells could be collected as dry swab samples for HPV detection.