Background Screening by cytology has been highly effective in reducing the incidence and mortality from squamous invasive cervical cancer (ICC), but the effectiveness is less established regarding non-squamous ICC and regarding women above screening ages and below 30 years of age. Cervical cancer still occurs despite the presence of an organised screening programme. A substantial proportion of screened women with ICC are reported to have had previous abnormal cytology findings. The significance of negative cytology with limited evaluation is not quite determined, the most effective management of women with low-grade abnormalities is controversial, and evaluation of long-term effect of different treatment methods is limited. Aims To identify possible areas of improvements in the prevention of cervical cancer by evaluating the effectiveness of the Swedish cervical screening programme, and by exploring risk factors for ICC in the cytological screening histories and in the management of women with abnormal cytology findings. Methods The screening histories of all ICC cases in Sweden 1999-2001 (n=1230) and of five population-based control women per case were reviewed, using data from the Swedish Cancer Registry, the national population register, the Swedish national cervical screening quality register, histopathological reports and questionnaires to clinicians. The risk of cervical cancer according to screening histories 0.5-6.5 years before cancer diagnosis was estimated as odds ratios (ORs) in logistic regression models with 95% confidence interval (CI) (Paper I). Risk related to different cytological reports was assessed in women below 67 years of age with cytology (n=572, n=3569) in Paper II. The initial follow-up of women with abnormal or unsatisfactory cytology reports (n=159, n=258) was evaluated in Paper III, and further investigation and treatment of abnormalities (n=143 cases, n=176 controls) in Paper IV. Results The cancer cases were above screening ages (31%), had not been screened according to recommendations (33%), had negative cytology (23%), or had previous positive screening tests (13%). No screening within the recommended interval increased the risk of squamous (OR 2.97, 95% CI 2.51-3.50) as well as non-squamous cancer (OR 1.59, 95% CI 1.20-2.11), and increased the risk in all ages. Negative cytology with partially obscuring factors and unsatisfactory cytology increased the risk of subsequent early stage ICC. All cytological abnormalities increased the risk of ICC, and women with glandular atypia or atypia in cells of uncertain origin carried a particularly high risk (OR 11.69, 95% CI 7.02-19.46). After a low-grade squamous abnormal smear finding, further investigation with biopsy was more effective than repeated cytology (OR 0.46, 95% CI 0.24-0.89). Lack of biopsy increased the risk in women with both low-grade and high-grade squamous abnormalities. Neither repeat cytology, nor biopsy, decreased the risk in women with glandular atypia or atypia in cells of uncertain origin. Treatment decreased the risk, even when the biopsy before treatment was negative or showed low-grade atypia only. Ablative therapy was less effective than excision and laser conisation was the most effective therapy. Conclusions Improved adherence to screening recommendations and including older women at increased risk in the programme would have significant cancer preventive gains. Women with negative cytology with limited evaluation and with unsatisfactory cytology may need further evaluation. Assessment with biopsy should be recommended for women with low-grade as well as high-grade squamous abnormalities. The diagnosing of precancer lesions and the identification of women in need of treatment warrant improvements, in particular in cases of glandular or “other” atypia in cytology. Treatment techniques need further evaluation.