Swedish civil servants, who reported considerable stress while working in larger government authorities, were randomized into either a work stress reduction intervention or a control group. Four work units constituted the intervention group (n = 94). and one unit formed the control group (n = 35). Participants in the two groups were of similar age and two thirds were women. Psychosocial, lifestyle, and physiological assessments were made before, during, and after the 8-month intervention period. A significant decrease in apolipoprotein B/apolipoprotein A I ratio (p < 0.05) occurred in the intervention group but not in the control group. No changes were seen in either group in smoking, eating. exercise, relative weight, or other lifestyle factors. which could explain the lipid profile changes. Stimulation from and autonomy over work increased in the intervention group (p < 0.01) hut remained the same in the control group. Results have implications for the efficacy of work stress interventions in decreasing coronary risk.