The way in which microbiological investigations are used in routine clinical practice and the value of such tests in directing antibiotic prescribing, was studied in adults admitted to hospital with a diagnosis of community-acquired pneumonia. One-hundred and twenty-two consecutive patients admitted to one teaching and one district general hospital were studied between April 1988 and March 1989. Blood cultures were performed in 81% of cases, sputum was examined in 45% and complete serological tests were performed in 28%. No causative pathogen was found in 74% of cases and results of microbial tests directed a change in antibiotic therapy in only 8% of cases. Routine microbial investigation of all adults admitted to hospital with community-acquired pneumonia is unhelpful and probably unnecessary. We suggest a strategy for microbial investigation linked to initial illness severity to replace the current haphazard approach.