Group medical audits should be considered as a substitute for, and even as a necessary first step in, the audit process ultimately applicable to individual patients. Group audits must take account of: 1. the proportionate distribution of individuals for each significant subgroup within a larger population, and 2. the clinical standards applicable to each of the specified subgroups. Data useful for audit standards are being developed at Yale by utilizing the opinions of practitioners as to the content of good medical care. Through the appropriate use of proportions and standards for various subgroups in a patient “mix,” overall weighted averages can be prepared for a mixture of patients regardless of whether the mix is based on single or multiple group characteristics.