This study investigated for the first time in the psychiatric literature the effect of parental age on age‐of‐onset (AO) in bipolar I disorder (BPI) in relation to proband sex and family history (FH) for major psychoses in a sample of 564 BPI probands. All probands, 72.68% of their first‐degree and 12.13% of their second‐degree relatives were directly interviewed. The FH‐method was used for all unavailable relatives. The diagnoses were made according to DSM‐IVTR. The impact of parental age on proband early/late AO was evaluated through logistic regression with the cut‐off for early AO determined through commingling analysis. We found evidence for a significant influence of increasing paternal age, and especially age ≥35 years, on AO of BPI disorder in the total sample (OR = 0.54, CI: 0.35–0.80), in the female subsample (OR = 0.44, CI: 0.25–0.78), in the sporadic subsample (OR = 0.64, CI: 0.38–0.95), and in the subsample with FH of recurrent unipolar major depression (Mdd‐RUP) (OR = 0.55, CI: 0.34–0.87). No significant effect of paternal age on disease AO was found in patients with FH of bipolar (BP), schizoaffective disorders (SA), or schizophrenia (SCZ), nor in males. Mean age was significantly higher in fathers of sporadic cases and of cases with FH of Mdd‐RUP than in fathers of cases with FH of BP/SA/SCZ (P = 0.011). Maternal age had no significant effect either in the total sample or in subsamples defined by proband sex or FH. In conclusion, in our sample increasing paternal age lowered the onset of BPI selectively, the effect being related to the female sex and FH‐type. © 2012 Wiley Periodicals, Inc.