All ski accident patients requiring an inhospital treatment at our institute from 1984 to 1992 (n = 2,053) were analyzed retrospectively. The incidence, pattern, and severity of the injuries, as well as the 30-day mortality comparing two time slots (1984 to 1988 and 1989 to 1992) were analyzed for the 361 cases classified as "serious" injuries according to the following definitions and groupings: group 1, multiple trauma (Injury Severity Score > or = 18) and patients with multiple fractures (n = 179; group 2, abdominal or thoracic single trauma (Abbreviated Injury Scale (AIS) score > or = 2, n = 58); and group 3, isolated head injuries (AIS score > or = 2, n = 124). Serious injuries were observed in 19% in the first period compared to 16% in the second period (not significant). Two hundred thirty-eight of 361 patients injured themselves by just falling, while 117 collided with some sort of obstacle like other skiers (45), trees or rocks (27), posts, pylons, barriers (20), and moving objects (25) such as piste machines, ski lifts, automobiles, and in one case a helicopter. Six skiers were caught on the ski run by an avalanche. Comparing the two time periods, trauma circumstances did not change significantly. The number of most severe head injuries (AIS score > or = 4) increased from 11.6 to 19.3% (p < 0.05). The overall mortality increased from 2 to 7% (p < 0.05). Group 2 had the best prognosis with no mortality, while group 3 was rated worst with 8%. Skiing remains a major source of serious trauma in winter resort areas.(ABSTRACT TRUNCATED AT 250 WORDS)