Diagnosis and treatment of 849 sore throat patients in primary care was studied with regard to the utilization and accuracy of a latex test for group A beta-hemolytic streptococci. 78% were tested, but 11% of the rapid test results were uncertain, and 15% conflicted with traditional culture results. The negative predictive value of the test was 98%, and the positive predictive value only 59%. 52% of all patients received antibacterials, the rapid test result influencing the treatment decisions clearly but not solely. The nurses utilized the rapid test more often than the physicians. The study stresses the difficulties inherent in introducing new tests in primary care and illustrates the realities of overutilization of the test and overmedication despite test results.