Although voriconazole (VRCZ) is metabolized to VRCZ N-oxide principally by CYP2C19, VRCZ clearance is affected by multiple factors. In this study, we investigated the relationship between the CYP2C19 phenotype using the VRCZ-to-VRCZ N-oxide plasma concentration ratio (VRCZ/N-oxide) and demographic and clinical characteristics of Japanese patients taking VRCZ.A total of 65 Japanese patients taking VRCZ for prophylaxis or treatment of fungal infection were enrolled in this study. Stepwise selection multiple linear regression analysis was performed to investigate the effect of factors on the VRCZ/N-oxide ratio.In patients not undergoing concurrent treatment with a drug influencing CYP2C19 activity (n = 54), the VRCZ/N-oxide ratio with definite thresholds for CYP2C19 genotypes, CYP2C19*1/*1, *1/*2 + *1/*3 + *2/*17, and *2/*2 + *2/*3, was specifically identified in patients taking VRCZ (<0.48, ≥0.48 < and <0.82 and ≥0.82). However, the VRCZ/N-oxide ratio could not be predicted based solely on the CYP2C19 genotype (R = 0.053). The route of VRCZ administration, C-reactive protein concentration determined on the same day as VRCZ plasma concentration measurement, CYP2C19 extensive metabolizer, and patient age were independent factors influencing the VRCZ/N-oxide ratio (R = 0.489, standardized regression coefficient = 0.385, 0.380, -0.231, and 0.231; P = 0.001, 0.001, 0.032, and 0.036, respectively).It is possible to comprehensively evaluate CYP2C19 activity using the actual measured value of the VRCZ/N-oxide ratio in patients taking VRCZ. The predictive performance of the VRCZ/N-oxide ratio was improved by including the route of administration, C-reactive protein level, and patient age in addition to the CYP2C19 genotype as predictive factors.