An increased high-risk human papillomavirus (hrHPV) viral load in cervical scrapings has been proposed as a determinant for high-grade cervical intraepithelial neoplasia (CIN) and cervical cancer (≥CIN 2), but data so far for HPV types different from HPV 16 are limited and inconsistent. In addition, a viral load threshold to distinguish hrHPV positive women without ≥CIN 2 still has not been defined. Here, we used baseline cervical scrapings of women with normal cytology participating in a large population-based cervical screening trial (i.e. POBASCAM) who were GP5+/6+-PCR positive for 4 common hrHPV types, i.e. HPV 16, 18, 31 or 33, as a reference to arbitrarily define various viral load thresholds (i.e. 25th, 33rd, 50th, 67th and 75th percentiles of the lowest viral load values) for distinguishing women having single infections with these types without high-grade CIN. Viral load assessment was performed by real time type-specific PCR. The viral load threshold values were subsequently validated on abnormal cervical scrapes of 162 women with underlying, histologically confirmed CIN lesions containing 1 of these 4 hrHPV types. All 59 women with CIN 3 had viral load levels that were higher than those of 33% of the women with normal cytology containing the respective hrHPV type detectable by GP5+/6+-PCR (i.e. higher than the 33rd percentile of viral load). By using this 33rd percentile viral load cut-off, sensitivity for CIN 3 of 100% (95% CI 93.9–100) was obtained. Hence, application of this viral load threshold would increase the specificity of HPV testing for HPV 16, 18, 31 and 33-associated prevalent CIN 3 without the cost of a marked reduction in sensitivity. In practice, on the basis of viral load analysis, a less aggressive management can be foreseen for 33% of the women with normal cytology participating in a population-based screening program who are GP5+/6+-PCR positive for HPV 16, 18, 31 or 33. © 2006 Wiley-Liss, Inc.