A low cost microcomputer-based medical record system was used to facilitate opportunistic research projects in the teaching clinic of the Chinese University of Hong Kong. An example of an actual clinical trial comparing the effectiveness of three methods of teaching breast self-examination to clinic patients was used and the process of setting up the research protocol, the sampling of patients, and the analysis of early results after one and a half years were described. It was found that the sampling bias in opportunistic research was that subjects selected were more likely to be older with more problems, and have a higher attendance rate at the clinic. The bias could be corrected during the analysis of data by stratifying the selected sample into groups of different attendance rates. Another potential sampling bias was due to non-compliance of both the patients and the clinical staff. However, it was found to be possible to obtain useful results from opportunistic research. It was easy to implement this type of research with the help of a computerized medical record system.