In 1974 several studies were conducted on the validity of medical data recorded and computerized at family medical centers affiliated with the University of Western Ontario. Fifty-nine encounters were observed. An average of 2.54 somatic, emotional, or social problems were dealt with per encounter. The residents recorded an average of 1.51 problems and the observers 2.45. This difference is highly statistically significant (p less than .001). There was no significant statistical differences among the observers. The many questions this study raises may have a bearing on medical education, medical audit, research, medical computer systems, and perhaps even on quality of care since problem-solving is based on problem identification. Further studies and evaluation are needed.