Continuity of care is defined and measured in terms of care received from a single physician, from several physicians practicing as a group and from physicians seen through referrals. All patients receiving tonsillectomy and adenoidectomy (T and As) in Manitoba for one year, as well as a group of similar individuals treated with respiratory illnesses but not having T and As, were studied. Several patient and physician characteristics were examined to determine their relationship to continuity of care. Continuity consistently increased with patient age, but appeared unrelated to several other variables. The links between continuity of care and quality of care were explored using process and outcome measures of quality. A cross-lagged panel analysis suggested no casual relationship between continuity and quality. These results contradict those of several other studies; they are important in that continuity of care is shown not to reduce one type of error of commission made by practitioners.