BackgroundThe optimal timing for therapeutic drug monitoring (TDM) of voriconazole in Asians, who have higher rates of poor metabolisers than non‐Asians, is unclear. This can cause unexpectedly high concentrations and delays in reaching steady‐state levels.ObjectivesTo determine the appropriate timing of TDM in Japanese patients receiving voriconazole.Patients/MethodsTrough levels (Cmin) were measured on days 3–5 (recommended timing, RT) and days 6–14 (delayed timing, DT) after starting voriconazole in patients receiving an appropriate dosage. Considering bioavailability, Cmin was only compared in patients receiving oral voriconazole.ResultsA total of 289 and 186 patients were included in the safety and pharmacokinetic analyses, respectively. There was a significant difference in Cmin measured no later than and after day 5 (3.59 ± 2.12 [RT] vs. 4.77 ± 3.88 μg/mL [DT], p = .023), whereas no significant difference was observed on cutoff day 6 (3.91 ± 2.60 vs. 4.40 ± 3.94 μg/mL, p = .465), suggesting that Cmin close to the steady‐state was achieved after day 5. DT causes a delay in achieving the therapeutic range. The hepatotoxicity rates were 21.5% and 36.8% in the RT and DT groups, respectively (p = .004); DT was an independent risk factor for hepatotoxicity.ConclusionAlthough steady‐state concentrations may not be achieved by day 5, early dose optimisation using RT can prevent hepatotoxicity in Japanese patients. TDM should be performed on days 3–5 to ensure safety. However, subsequent TDM may be necessary due to a possible further increase in Cmin.