We conducted a randomized controlled trial (RCT) to examine the effects of mailed advice on reducing psychological distress, blood pressure, serum lipids, and sick leave of workers employed in a manufacturing plant in Japan. Those who indicated higher psychological distress (defined as having GHQ scores of three or greater) in the baseline questionnaire survey (n = 226) were randomly assigned to an intervention group or a control group. Individualized letters were sent to the subjects of the intervention group, informing them of their stress levels and recommending an improvement in daily habits and other behaviors to reduce stress. Eighty-one and 77 subjects in the intervention and control groups, respectively, responded to the one-year follow-up survey. No significant intervention effect was observed for the GHQ scores, blood pressure, serum lipids, or sick leave (p > 0.05). The intervention effect was marginally significant for changes in regular breakfasts and daily alcohol consumption (p = 0.09). The intervention effect was marginally significant for the GHQ scores among those who initially did not eat breakfast regularly (p = 0.06). The study suggests that only sending mailed advice is not an effective measure for worksite stress reduction. Mailed advice which focuses on a particular subgroup (e.g., those who do not eat breakfast regularly) may be more effective.