In order to detect possible relationships between antisocial behaviour and the incidence of "sudden violent death" in young people, information relating to mortality in antisocial Swedish adolescents has been traced and compiled. A register was drawn up covering those young persons (1,056; 832 boys and 224 girls; mean age 16 years) who were admitted to Swedish probationary schools during the period 1 January - 31 December 1967. Using the registers of immigration and emigration, and causes of death kept by SCB (Statistiska Centralbyrån), mortality occurring between 1 January 1967 - 31 December 1985 was tabulated. One hundred and ten boys (13%) and 22 girls (10%) had died. The deaths had occurred at a rate of approximately seven new deaths per observation year, the youngest being still in their teens when they died. For comparison, the criteria set up by insurance companies for life insurance premiums are based on a death expectancy for healthy Swedish boys and girls in the age groups corresponding to the subjects under observation of 1.2-3.1% for boys and 1.1-2.6% for girls. Eighty-eight percent of the dead boys and 77% of the dead girls had died "sudden violent deaths" - accidents, suicides, death from uncertain causes, murder/manslaughter, or alcohol/drug abuse. For both sexes, death from uncertain causes and suicides were the most frequent single causes of death. Death as a direct result of alcohol/drug abuse occurred only in boys. The results give support to the assumption that a link exists between childhood environment, the development of antisocial behaviour/mental insufficiency and a "sudden violent death" at an early age.(ABSTRACT TRUNCATED AT 250 WORDS)