The possible relationship between a number of biochemical parameters and measures of pain and depression was studied in chronic pain patients without a major depression. In a double-blind crossover study, patients were treated with amitriptyline combined with a low dose of flupentixol or placebo. We investigated whether pretreatment biochemical values correlated with initial data on pain and/or depression, or whether they had predictive value for treatment outcome. We also studied systematically the effect of both treatment regimes on the biochemical parameters themselves and their relation to the plasma levels of amitriptyline. From our results, the possible involvement of the serotonin system in somatoform pain disorder is confirmed and no direct relation with the noradrenergic system could be inferred. The lack of involvement of a number of putative, depression-related, biochemical parameters suggests that affective disorders and pain syndromes do not share all mechanisms in common.