Objectives: To explore the reasons for the 40-fold variation in diagnostic testing for genital Chlamydia trachomatis by general practices. Methods: A qualitative study with focus groups. We randomly selected urban and rural high and low testing practices served by Bristol, Hereford, and Gloucester microbiology laboratories. Open questions were asked about the investigation of C trachomatis in men and women in different clinical contexts. Results: The high and low testing practices did not differ in their age/sex make-up or by deprivation indices. There were major differences between high and low chlamydia testing practices. Low testing practices knew very little about the epidemiology and presentation of genital chlamydia infection and did not consider it in their differential diagnosis of genitourinary symptoms until patients had consulted several times. Low testers were less aware that chlamydia was usually asymptomatic, thought it was an inner city problem, and had poor knowledge of how to take diagnostic specimens. High testing practices either had a general practitioner with an interest in sexual health or a practice nurse who had completed specialist training in family planning. High testing practices were more cognizant of the symptoms and signs of chlamydia and always considered it in their differential diagnosis of genitourinary symptoms, including patients attending family planning clinics. Conclusions: Any programme to increase chlamydia testing in primary care must be accompanied by an education and awareness programme especially targeted at low testing practices. This will need to include information about the benefits of testing and who, when, and how to test.