BACKGROUND A screening programme to detect polyps or early carcinoma would significantly reduce the mortality and morbidity of colorectal cancer (CRC). The aims of the present study were to evaluate: (i) the feasibility of training general practitioners in flexible sigmoidoscopy (FS) for CRC screening; (ii) the acceptability of screening by faecal occult blood testing (FOBT) and FS in asymptomatic standard risk Australians aged over 50 years; and (iii) the yield of such screening. METHODS Subjects were recruited by general practitioner (GP) referral, newspaper advertisement or by a direct approach to retirement villages. Participants were mailed a FOBT kit and a prescreening questionnaire. Flexible sigmoidoscopy was performed by a GP supervised by an experienced endoscopist. Subjects then completed a second questionnaire. General practitioners were assessed after 50 unassisted procedures. RESULTS A total of 264 individuals contacted the study coordinator; 169 were screened. Screening was accepted well by the participants. Fifteen per cent of subjects had polyps and 4% had a positive FOBT. Training in FS was adversely affected by the availability of resources. Three GPs completed 50 unassisted procedures over a 15-month period, but none were able to reliably assess the distal bowel. CONCLUSIONS Although the three trainees and their supervisors did not consider that the GPs were adequately trained after 50 unassisted procedures, training was adversely affected by limited resources within the Victorian public hospital system. Screening by FOBT and FS was considered to be acceptable by the patients undergoing these procedures. Existing facilities are not adequate if GPs are to be trained in FS as part of a national CRC screening program.