Cervical cancer screening in Belgium is mainly opportunistic with periodic attempts to organize it according to the European guidelines. In the province of Limburg (north-east Belgium), a cervical cytology registry was set up in collaboration with local cytopathological laboratories, provincial health authorities and the Limburg Cancer Registry. Laboratories regularly communicated coded results of Pap smears from women residents in Limburg to the provincial cytological registry. All individual records contained a virtually unique identifying code allowing the study of longitudinal histories and linkage with the cancer registry. The screening coverage (percentage of women with a Pap smear in a defined period), the prevalence and incidence of squamous intraepithelial lesions and changes over time and geographical area were evaluated using a database of more than 600 000 Pap smear interpretations between 1996 and 2005. In 2000, 47% of women aged between 25 and 64 years had at least one Pap smear recorded in the last 3 years, and the average number of smears screened for each woman was 1.5. On account of incomplete registration, the coverage was 8% lower than estimated from health insurance data. The modal screening interval was in the range of 12-14 months; the first quartile, median and third quartile were 355, 440 and 676 days, respectively. Over the 10 years, 82% of the target population had their Pap smear recorded. The prevalence of cytological abnormalities was 3.7% (atypical squamous cells of undetermined significance: 2.2%, atypical glandular cells of undetermined significance: 0.1%, low-grade squamous intraepithelial lesion: 1.1%, high-grade squamous intraepithelial lesion or more serious lesions: 0.4%), but varied substantially among laboratories. The prevalence of atypical squamous cells of undetermined significance and low-grade squamous intraepithelial lesion increased significantly over time. Pathologists from Limburg have pioneered cytology registration in the context of opportunistic screening in Belgium. The estimated screening coverage was slightly underestimated. Obligatory registration, use of a uniform terminology and linkage with population and follow-up data are needed to give a cytology registry its full role in an organized screening programme.