The rate of congenital malformations after first-trimester exposure to itraconazole was four times higher when ascertained retrospectively than prospectively (13.0 vs 3.2%, p = 0.006). Reporting bias in retrospective studies should be acknowledged in interpretation of such data.