The use that 30 general practitioners in four group practices made of open access laboratory and radiological facilities was studied for one year. We were particularly interested in whether general practitioners hoped to exclude rather than confirm abnormality when requesting investigations. All but two of the general practitioners studied used investigations to exclude abnormality to a greater extent than to confirm it. The rate at which the practices investigated patients and the number of investigations requested were appreciably different and were different for individual general practitioners, part time general practitioners requesting more investigations than trainees and full time general practitioners. Haematological investigations accounted for over 30% of requests for investigations in all but one practice, biochemical investigations being requested as often as bacteriological investigations in two of the four practices. The ratio of expected to unexpected results varied among general practitioners; no general practitioner had more unexpected results and the range of ratios was similar for full time and part time general practitioners and for trainees. The proportion of patients with abnormalities uncovered by each practice increased disproportionately as the use of investigations increased, supporting a higher rather than lower rate of investigation among general practitioners. Compared with the results of other studies the use of the radiological facilities available was low. X ray examinations of the skeleton were requested more than chest and contrast media examinations by three of the four practices. Most x ray examinations were used to exclude rather than confirm abnormality by all the practices, with over 85% of results confirming the general practitioner's initial diagnosis.