The acidic, sandy soils in the southeastern US are phosphorus (P) limited for forest production and are commonly fertilized with P. There is no P retention capacity in the A horizon. However, microbial biomass may immobilize and retain P fertilizer before it is leached below seedling rooting depth making P fertilization more efficient. An accurate estimation of microbial P is dependent on measuring the K-p factor in the fumigation-extraction method. The overall purpose of this study was to examine the fumigation-extraction method for microbial P in acidic, forested, sandy soils. The three objective were: to determine which extractant was the most useful extracting microbial P by comparing the standard basic extractant, 0.5 M NaHCO3 at pH 8.5, against several acidic and oxalate extractants; to evaluate whether soil water potential influenced the K-p factor; and to test whether the K-p factor differed by soil horizon within the profile of a representative Flatwoods Spodosol. Three millimolar oxalate was determined to be the preferred extractant due to its efficient removal of microbial P and ease of analysis. The K-P factor was dependent on soil water potential and horizon. The range in K-p at different water potentials using 3 mM oxalate was from 0.31 to 0.67 in the A horizon, 0.48 to 0.91 in the E horizon, and 0.22 to 0.45 in Bh horizon. The highest K-P factors tended to be at water potentials near saturation and under the driest condition. Differences in K-p were attributed to the influence that water potential and soil horizon had on microbial assemblages and diversity. Using a literature value of K-p, instead of measuring K-p directly, caused an overestimate of -7 to 63% in the A horizon, 63-160% in the E horizon and 7-32% in the Bh horizon. The best estimate of microbial P required that K-P be evaluated for specific soil conditions. (C) 2004 Elsevier Ltd. All rights reserved.