Phosphorus (P) fractions are commonly assessed on dried soils in environmental and long-term field research, because this is the most convenient method of storing large quantities of samples before analysis. However sample pre-treatment may seriously affect the results, especially those regarding the organic P fractions. The objective of the present study was to assess the effect of soil drying under laboratory conditions on the P status of soils with contrasting physico-chemical and microbial properties. We used soils from a ( Pinus pinaster) forest in southwest France sampled from different soil depths and different site classes featuring a gradient of soil organic matter (SOM), and aluminium (Al) and iron (Fe) oxides. Total P, total organic and inorganic P, water soluble phosphate ions (iP), microbial P and the potentially mineralizable P in 'dead' organic matter were determined on fresh, moist soils maintained at 4°C and on the same soils, oven-dried at 60°C. Laboratory-drying resulted in a general decrease in total organic P, which was positively correlated with microbial P and SOM content. These drying-induced changes in total organic P were associated with a concomitant increase in total inorganic P and water soluble iP, which were dramatic in the organic forest floor samples (on average +2965% (*31 increase) in total inorganic P, and +11880% (*123 increase) in water soluble iP). The changes in water soluble iP were correlated with the SOM: Al and Fe oxides ratio and showed little effect of soil drying in deeper soil layers with a low SOM and high oxide contents. Our results indicated that soil drying mainly resulted in the lysis of the microbial pool, and to a much lesser extent, in the physical disruption and mineralization of the SOM. We conclude that soil drying before analyses can generate a significant bias in the evaluation of soil P stocks and fluxes, especially in SOM-rich layers. We recommend that fresh soil samples should be used as far as possible to determine soil P status.