Several lines of evidence implicate dopa decarboxylase (DDC) with schizophrenia. By analysis of two putative functional DDC variants in 173 schizophrenic patients and 204 controls we tested the hypotheses that DDC is involved in: (1) predisposition to schizophrenia; and (2) modulation of age at disease onset. No association was observed with schizophrenia as a whole, whereas an association between DDC genotypes and age at disease onset was suggested in males (P = 0.03). This association was most pronounced in relation to genotypes of haplotypes comprising both variants, suggesting an additive model where one variant mediates early and the other late onset. Accordingly, the haplotype-based genotypes could be assigned into three groups by their possible relative effect on age at onset: an "early", "neutral" and "late" group. Dividing the male schizophrenics into four groups with increasing age at onset, the "early" genotypes were seen to decrease in frequency from 51.5% to 16.7% while the "late" genotypes increased from 12.1% to 33.3% (P = 0.02). The difference in mean age at onset between male patients with "early" genotypes vs patients with "late" genotypes was close to 5 years (95% CI: 0.7-8.8). Thus, DDC may possibly act as a modulator of age at onset in male schizophrenics.