This study aimed to validate the fumigation-extraction method for measuring microbial biomass P in acid soils. Extractions with the Olsen (0.5 M NaHCO3, pH 8.5) and Bray-1 (0.03 M NH4F–0.025 M HCl) extractants at two soil:solution ratios (1 : 20 and 1 : 4, w/v) were compared using eight acid soils (pH 3.6–5.9). The data indicated that the flushes (increases following CHCl3-fumigation) of total P (Pt) and inorganic P (Pi) determined by Olsen extraction provided little useful information for estimating the amount of microbial biomass P in the soils. Using the Bray-1 extractant at a soil:solution ratio of 1 : 4, and analysing Pi instead of Pt, improves the reproducibility (statistical significance and CV) of the P flush in these soils. In all the approaches studied, the Pi flush determined using the Bray-1 extractant at 1 : 4 provided the best estimate of soil microbial biomass P. Furthermore, the recovery of cultured bacterial and fungal biomass P added to the soils and extracted using the Bray-1 extractant at 1 : 4 was relatively constant (24.1–36.7% and 15.7–25.7%, respectively) with only one exception, and showed no relationship with soil pH, indicating that it behaved differently from added Pi (recovery decreased from 86% at pH 4.6 to 13% at pH 3.6). Thus, correcting for the incomplete recovery of biomass P using added Pi is inappropriate for acid soils. Although microbial biomass P in soil is generally estimated using the Pi flush and a conversion factor (k P) of 0.4, more reliable estimates require that k P values are best determined independently for each soil.