INTRODUCTION:The aim of this study was to assess population-based changes in incidence, treatment, and in short- and long-term survival of patients with acute respiratory distress syndrome (ARDS) over 23 years. MATERIALS AND METHODS:Analysis of all patients in Iceland who fulfilled the consensus criteria for ARDS in 1988-2010. Demographic variables, Acute Physiology and Chronic Health Evaluation II (APACHE II) scores and ventilation parameters were collected from hospital charts. RESULTS:The age-standardised incidence of ARDS during the study period was 7.2 cases per 100,000 person-years and was increased by 0.2 cases per year (P < 0.001). The most common causes of ARDS were pneumonia (29%) and sepsis (29%). The use of pressure-controlled ventilation became almost dominant from 1993. The peak inspiratory pressure (PIP) has significantly decreased (-0.5 cmH(2) O/year), but the peak end-expiratory pressure (PEEP) has increased (0.1 cmH(2) O/year) during the study period. The hospital mortality decreased by 1% per year (P = 0.03) during the study period, from 50% in 1988-1992 to 33% in 2006-2010. A multivariable logistic regression model revealed that higher age and APACHE II score increased the odds of hospital mortality, while a higher calendar year of diagnosis reduced the odds of mortality. This was unchanged when dominant respiratory treatment, PIP and PEEP were added to the model. The 10-year survival of ARDS survivors was 68% compared with 90% survival of a reference population (P < 0.001). CONCLUSION:The incidence of ARDS has almost doubled, but hospital mortality has decreased during the 23 years of observation. The 10-year survival of ARDS survivors is poor compared with the reference population.