The non-participation to cervical screening is the major determinant in the risk of mortality due to cervical cancer. In France, around 40% of women do not participate to regular screening. The cultural or economic barriers for performing screening by Pap test are numerous; one of the most frequent is the refusal of gynaecological examination. A persistent HPV(HR) infection is a necessary factor for developing cervical cancer. The HPV(HR) testing has a high sensibility to detect high grade cervical intra-epithelial neoplasia (CIN 2-3) and a satisfactory specificity after 30-35 years old. The principal objective of this study was to compare the participation rates in women 35-69 years old who did not perform a Pap test after a first individual invitation, either when an HPV(HR) auto-test was offered to be performed at home or a second invitation to Pap test was sent. We also evaluated the quality of the two tests, the positive results obtained by age groups and the following histological type of lesions diagnosed in the women with positive results. The study included 9,334 women, 35-69 years old, who did not realized a Pap-test during the 2 previous years and who did not respond at a first individual invitation. These non-responders were randomized into two groups: one group (n=4,934) received a second individual invitation and the other (n=4,400) an offer of receiving and performing an HPV auto-test at home. In women 35-69 years the participation to the second invitation to Pap test was significantly lower (7.2%) than the participation to auto-test (26.4%) with P<0.001. The quality of the two tests was satisfactory; the auto-test was not altered by the postage to laboratory (non interpretable rate=1.4% [CI at 95%=0.65%; 2.15%]. From the 311 Pap tests done, 5.5% (17) were classified "abnormal" (nine ASCUS, one high grade and seven low grades). The follow up of 13 women out of 17 confirmed the diagnosis for 1 case of CIN2 and 2 cases of CIN3, 4 women are lost of follow up after 6 months. From the 939 HPV(HR) tests done, 6.2% (58) were positive. Such positivity rate was not influenced by age. Out of the 58 positive HPV(HR) cases, 27 only were of the 16 genotype (46.5% [CI 95%=33.7%; 59.3%]). This law rate is a consequence of an inversion of the ratio HPV 16 versus other types in women 60 years old and over. In this group, the follow-up of 36 women diagnosed five cases of CIN1, one of CIN2 and four of CIN3; 22 patients are lost of follow up at 6 months. Globally, in the studied population, an individual recall for pap test allowed to diagnose and treat 3 high grade lesions (7‰) and the dispatching of an auto test allowed the diagnosis and treatment of five high grade lesions (1,4‰), this difference is significant (P=0.02; OR=0.25 [0.05; 0.97]). The HPV(HR) auto-test seems to be better accepted than the Pap test in the 35-69 years old women previously non-responders to individual invitation, and the quality of the test is satisfactory. Such a test can be proposed to the 35-69 years old non-participant to Pap test to increase the coverage for cervical screening, if the rates of diagnostic examinations performed in case of an HPV(HR) positive is sufficiently high.