Mycoplasma pneumoniae is a frequent cause of community-acquired respiratory infections in children and adults. Although the organism is felt to be the most frequent 'atypical' pathogen responsible for community-acquired pneumonia in adults, the prevalence of M. pneumoniae varies greatly from study to study, depending on the population and the diagnostic methods used. Recent studies have found the prevalence of M. pneumoniae in adults with pneumonia to range from 1.9 to over 30%. M. pneumoniae is also a frequent cause of outbreaks of respiratory disease in institutional settings. However, the diagnosis of M. pneumoniae infection is hampered by the lack of standardized, rapid, specific methods. This problem was illustrated by the results of an investigation of an outbreak of M. pneumoniae infection in a federal training facility. Accurate diagnosis required a combination of polymerase chain reaction and serology, as IgM antibodies were not present early in the course of the infection in many patients. Several papers evaluating various serological and polymerase chain reaction assays were published during the period of this review. An assessment of the actual performance of these tests was also hampered by the lack of standardized comparative methods. M. pneumoniae is susceptible in vitro to macrolides, tetracyclines and quinolone antibiotics; however, data are limited on the microbiological efficacy of these agents. Several pneumonia treatment studies were published during this period, practically all of them based the diagnosis of M. pneumoniae infection on serology; different methods and criteria were used in each study, and thus the microbiological efficacy could not be assessed. The Infectious Disease Society of America recently stated in their revised Practice Guidelines for the Management of Community-Acquired Pneumonia in Adults that, as there were no diagnostic tests available that reliably and rapidly detect M. pneumoniae, therapy must usually be empirical.