Psychoendocrine and behavioural effects elicited by viewing an exciting and partly violent movie were studied in 12-year-old boys ( n = 42). The boys were divided into three groups. Group A ( n = 13) consisted of boys scoring below the median value on the Sennton neurotic scale for vegetative anxiety symptoms. Group B ( n = 13) consisted of boys scoring above the median value on the Sennton neurotic scale. None of these boys had ever visited a psychiatric treatment centre, and they all came from the same school. A third group, Group C ( n = 16), consisted of boys who, at least once, had visited an out-patient clinic at a child psychiatric treatment centre for neurotic problems. These boys attended different schools in the hospital catchment area. Group C had the same mean score on the Sennton neurotic scale as did Group B. By having two groups of boys scoring both high on the Sennton neurotic scale, one of which knew the other boys and one that did not, we were able to study possible effects of social support on psychobiologic reactions in neurotic boys. The boys' reactions to the movie were assessed by: (1) self-ratings; (2) direct observations of the boys by a team of trained child psychologists; and (3) by analysis of urinary output of adrenaline, noradrenaline and cortisol. Group B rated their experience the strongest, viewing the movie, while Groups A and C rated themselves as rather unaffected by it. Group C was rated less verbally active, both during the control and the film periods, than were the other two groups. Group A was most motorically active during the film period. There were no significant differences during the control period in endocrine variables. During the film period Group C excerted significantly more noradrenaline than did Group A, and significantly more noradrenaline than did Group B. Intra-group differences in response to the movie were also assessed. Adrenaline excretion increased significantly only in Group C during the movie, compared with the control period. Furthermore, excretion of noradrenaline decreased significantly during the film period in Groups A and B. The A and B groups preferred to review scenes representing danger, while Group C preferred scenes representing security. It is suggested that psychosocial factors, e.g. novelty to the environment and lack of friends, caused the differences between Group C and the other two groups. Group C, which scored the same as Group B on Sennton neurotic scale, did not have access to social support in the form of friends during the movie. They were left alone with their anxiety, with little ability to express it in a group of unknown boys. Group B appeared to have coped in a more ‘open way’, expressing their emotions during the film period, possibly explaining this group's lower level of physiological arousal. The psychological vulnerability in Group C is further stressed by this group's desire to review scenes depicting security. It suggested that access to social support, here operationally defined as being together with friends, is an important intervening variable in predicting psychobiological effects of viewing exciting and violent motion pictures.