Most previous reports of Rocky Mountain spotted fever (RMSF) have included cases either not laboratory confirmed or confirmed by relatively weak diagnostic criteria. In the present study detailed epidemiological, clinical, and laboratory features of 262 confirmed or highly probable cases of RMSF reported from six states from 1977 to 1980 were analyzed. This analysis revealed that early clinical diagnosis of RMSF is difficult because the illness may have a gradual or an abrupt onset, the symptoms and signs may be unusual in timing or frequency, and the clinical appearance may vary depending on such factors as age and location of residence. RMSF was diagnosed later in those who died than in survivors, primarily because of atypical initial symptoms and the late onset of rash. RMSF should be considered in any individual who, during the spring and summer, has been in RMSF-endemic areas and develops a fever, regardless of the absence of rash or history of tick exposure.