Although it has been argued that those with lower levels of serum cholesterol are likely to be depressive, the findings are inconsistent. The present study attempted to clarify the relationship between major depression and serum total cholesterol in a working population. Subjects were 987 Japanese men working at an institute, aged 20 to 64 years. In addition to blood examinations and physical measurements, clinical structured interviews of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) were used to detect major depression. The prevalence of major depression was higher in the hypercholesterolemics (serum total cholesterol levels ≥ 5.69 mmol/liter) than in the normocholesterolemics (3.10-5.69 mmol/liter) (6.1% vs 1.8%, p < 0.05). Notably, there was no case with major depression among the hypocholesterolemics (< 3.10 mmol/liter). Through a multiple regression analysis, serum total cholesterol levels were positively predicted by the following four variables: major depression, age, body mass index, and skipping breakfast (all p < 0.01). Concerning those diagnosed with major depression, serum total cholesterol levels remained higher in the following year (p < 0.05), comparing to those without such diagnosis. Therefore, depression is associated with higher serum cholesterol levels in a population of Japanese male workers. The irregularity of eating behavior may be one of the factors mediating high serum cholesterol levels and major depression.