We conducted a survey to determine the accuracy of the clinical diagnosis of measles in Zimbabwe. Between December 1996 and February 1997, we collected blood samples and clinical and demographic information from a sample of 105 children with a clinical diagnosis of measles. A clinical case of measles was defined as a person with a history of fever, rash for three or more days, and either cough, coryza, or conjunctivitis. A laboratory-confirmed case of measles or rubella had IgM antibodies against measles virus or rubella virus respectively. A total of 91% of children met the clinical case definition. Among those who met the clinical case definition for measles, 72% were IgM-positive for measles virus only, 23% were IgM-positive for rubella virus only, 3% were IgM-positive for both measles and rubella viruses, and 2% were IgM-negative for both viruses. This study demonstrates the importance of considering selective laboratory confirmation of measles in periods of high disease incidence when the effectiveness of the vaccine is questioned.