In 277 patients admitted to hospital for community-acquired pneumonia (CAP) an aetiologic diagnosis was established in 68% with S. pneumoniae being the predominating agent. Four percent of the patients (12/277) died during their hospital stay, and only one of these patients was below 60 years of age. On admission, the most important factor, independently associated with fatal disease was a low serum albumin concentration, which was also a negative prognostic factor for the course of the survivors. In patients admitted to hospital for CAP, the finding of a low serum albumin level should therefore lead to intensified observation and treatment. Of 241 patients discharged after treatment for CAP, 50 patients were readmitted to hospital with recurrence of pneumonia during a 31 month follow-up period. This pneumonia incidence rate was more than five times that in a control population. Fifty-one of the patients (21%) died during follow-up, with 13 (25%) of the deaths directly associated with pneumonia. Systemic treatment with corticosteroids was associated with a higher risk of recurrence of pneumonia and death, while airway colonisation with Gram-negative enteric bacteria and a serum albumin below 30 g/l during hospital treatment of the initial pneumonia were associated with death from pneumonia after discharge. In 97 middle-aged and elderly patients admitted to hospital for CAP, malnutrition reflected by low triceps skinfold (TSF) and body mass index (BMI) values was associated with death during a six-month follow-up period, as was severity of disease on admission classified according to acute physiology and chronic health evaluation (APACHE II). Admission serum concentrations of orosomucoid and alpha-1-antitrypsin were most closely correlated with in-hospital morbidity measured as days spent in hospital and duration of fever. The risk of readmission within six months of discharge was higher in patients with high admission levels of APACHE II and TSF. Measurement of serum concentrations of alpha-1-antitrypsin and orosomucoid on admission should be considered in order to better predict hospital morbidity in these patients. Measurements of APACHE II and TSF on admission may give additional prognostic information on the interval from admission to six months after discharge. On admission 64% of the patients were hypoalbuminaemic, but only 6-10% were so at follow-up visits. Admission serum albumin concentration correlated negatively with investigated acute-phase proteins, and positively with other serum transport proteins, but no association with investigated nutritional measurements was found. The main reason for depressed serum albumin in elderly patients with pneumonia thus seems to be not malnutrition, but the inflammatory reaction per se. In 203 hospital-treated patients with CAP, the diagnostic and prognostic value of admission serum levels of interleukin-6 (IL-6) and C-reactive protein was investigated. The highest levels of IL-6 and CRP were found in patients with pneumococcal pneumonia, especially when bacteraemic. Patients with high IL-6- or CRP levels had longer duration of fever, longer hospital stay, and fewer had recovered clinically or radiographically at follow-up eight weeks after discharge. A high IL-6, but not a high CRP, also seemed to be associated with a higher mortality. The type-specific antibody responses to six pneumococcal capsular polysaccharide antigens included in the 23-valent vaccine as well as antibodies against the vaccine were measured by use of an enzyme-linked immunosorbent assay in 65 middle-aged and elderly individuals treated in hospital for pneumonia eight weeks prior to vaccination. The antibody concentrations before and after the vaccination were comparable with those in a vaccinated age-matched control group who had not recently been treated for pneumonia...