Abstract Therapeutic drug monitoring (TDM) of antifungal triazole was recommended, except for isavuconazole (ISA) whose target trough concentrations ( C min ) need to be specified. Concerning posaconazole (POS), tablet formulation results in higher exposure but no upper C min threshold has been yet proposed. We aimed to investigate the pharmacokinetic–pharmacodynamic relationship of POS and ISA, using a bioassay approach as surrogate marker of antifungal activity, in order to refine the therapeutic C min of both antifungals. A bioassay using a cellulose disk diffusion method was performed to determine the growth inhibition zone (GIZ) of POS and ISA on Aspergillus fumigatus and Candida parapsilosis (ISA only). GIZs of plasma from patients undergoing TDM for POS ( n = 136) or ISA ( n = 40) were determined. GIZs of plasma patients and antifungal C min were highly correlated for ISA ( A. fumigatus : ρ = 0.942, P < 0.0001; C. parapsilosis : ρ = 0.949, P < 0.0001) and POS (ρ = 0.922, P < 0.0001), and these relationships were represented with a Michaelis–Menten model. Based on this modeling, the recommended thresholds of 0.7, 1, and 1.25 mg/L for the POS C min corresponded to 50.1, 55.2, and 59.1% of the maximal GIZ, respectively. We propose an upper threshold of 4.8 mg/L for the POS C min and a lower threshold of 2.0 mg/L for the C min of ISA, as they respectively corresponded to concentrations leading to 90% and 50% of the maximal GIZ on A. fumigatus . The determination of antifungal activity using this bioassay allowed refining target C min of POS and ISA, especially the upper threshold of POS (4.8 mg/L) and the lower threshold of ISA (2.0 mg/L).