In 71 males who survived acute meningococcal disease 3 to 15 years ago at an age of about 20, associations between acute clinical conditions (including a few pre- and post-admission variables) and late sequelae have been studied. There was a higher rate of sequelae symptoms (mainly light neurological and mental disturbances) among survivors from meningitis (76%) than among those who had had both meningitis and septicemia (58%) or pure septicemia (50%). Twenty percent of control persons experienced such symptoms. "Changed Life" because of serious educational and working problems followed in 29% of the meningitis cases and 70% of the septicemia cases. Most of the clinical and laboratory factors separately examined were not significantly correlated to the sequelae rates. However, less than 2.5 mmol/l glucose in the cerebrospinal fluid (CSF) on admission (p less than 0.01), more than 1000 X 10(6) white blood cells per 1 in the cerebrospinal fluid (p less than 0.05), fever for more than 8 days (p less than 0.05), and probable cerebral symptoms the first week (p less than 0.05), were all positively correlated to a high rate of late sequelae. Well documented early sequelae correlated with serious late sequelae (p = 0.05). No conspicuous associations between acute antibiotic treatment and late sequelae were found. A combination of CSF glucose, blood thrombocytes, and cells in CSF on admission yielded a multiple regression score which seems to be a moderately reliable predictor of sequelae (R = 0.46). Hospital treatment should both aim at avoiding death and escaping residual effects. Because many prognostic factors for sequelae on admission are different from those for lethality, scoring for sequelae may be helpful in such secondary prevention of sequelae. Early standardized registration of sequelae may also be of value in tertiary prevention.