As part of the research into the effect in the consultation of the use of a computer to prompt opportunistic preventive care a valid, objective, and practical measure of the consultation process was required. After a review of the alternative methods the Time Interval Medical Event Recorder (Timer) was developed, its reliability tested, and applied to 93 control consultations and 49 computer assisted consultations. Timer records, every five seconds, four consultation events: the problems being dealt with, the physical activity, the verbal activity, and the secondary tasks being attempted. Timer showed that control consultations lasted an average of 6 minutes 58 seconds. The doctors spent 35% of their time on administration, and patients and doctors were both conversational for just 33% of the consultation. Giving information was the most common verbal activity (48% of the duration of the consultation) with silence accounting for 21% of the time. When the computer was used the average consultation was longer, at 7 minutes and 46 seconds. The doctor's contribution to the consultation appeared to have increased. Patient centred speech fell from 36% in controls to 28% of the duration of the consultation when the computer was used, while doctor centred speech rose from 30% to 34.5%. Secondary tasks (exploring patient concepts, education, management sharing, and prevention) were attempted during 28% of the control consultations and 40% of the computer consultations. This was accounted for by the increase in prevention (p less than 0.001). Timer is a reliable and practical tool for researching the consultation, and though it has shown validity in detecting differences between consultations that use a computer and those that do not, further applications are required to establish its full value.