to analyse the prevalence of cervical chlamydia infection and its determinants in an Italian population of women attending outpatients services for contraceptive counselling or routine gynaecological examination. between November 1989 and November 1990 we conducted a cross-sectional study on the prevalence of cervical Chlamydia trachomatis infection among women attending the outpatients service of seven university clinics in Northern (three centres), Central (three centres) and Southern (one centre) Italy. Eligible for the study were subjects with symptomatic low gynaecological tract infection (a total of 2071 women), a history of recurrent abortions (two or more miscarriages and no livebirth (416 subjects)), or sterility (371 subjects), plus a sample of asymptomatic women observed for contraceptive counselling or routine gynaecological examination identified on randomly selected days at the participating centres (1321 subjects). During the gynaecological consultation women were asked about their general characteristics, reproductive history, contraceptive and sexual habits, and history of sexually transmitted diseases (STD) using a standard questionnaire. An endocervical specimen was obtained with a plastic swab. The direct smear immunofluorescent antibody test (IFA test) was used to detect chlamydia antigens. out of the 2071 women with genital infection, 104 (5.0%) had cervical chlamydia infection; the corresponding percentages were 4.6 (19/416), 5.4 (20/371) and 3.9 (51/1321) respectively in women with recurrent abortions, sterility and in asymptomatic subjects. The risk of chlamydia infection was higher in women reporting a history of STD: in comparison with those without a history of STD, the relative risk of chlamydia infection was 1.4 (95% confidence interval, CI, 1.0-2.0). Among women reporting current use of a contraceptive method the risk of cervical chlamydia infection was lower in current users of barrier methods; in comparison with oral contraceptive users, the RR was 0.4 (95% CI, 0.2-0.8) in barrier methods users and 0.5 (95% CI, 0.2-1.1) in intrauterine device or other methods users. No consistent relationship emerged with age, reproductive history or number of sexual partners over the last 12 months. in this Italian population the frequency of cervical chlamydia infection appeared to be lower than in other selected groups from Northern European and American countries. Users of barrier contraception methods were at reduced risk of infection.