The effect of different extracting solutions at two soil/solution ratios (1:10, 1:25) and shaking periods (5 min and 16 h) on the availability of residual and recently applied Pin some oxisols was examined by comparing the North Carolina method with the Bray No.1 and Olsen procedures. Correlation coefficients between available P and P extracted by the three methods were highly significant for all the soils. When the soils were grouped according to texture (sands and clays), coefficients were 0·93 in the sandy soil group as against 0·58 in the clayey soils for a 90 day incubation period. In the samples where no P was applied the North Carolina method extracted 30 per cent more P for the 16 h shaking period than for the 5 min period, whereas P reabsorption occurred in the other methods (ratio 1:10). In the 250 p.p.m. P treatment for the 90 day incu-bation period, the amount of P extracted by the North Carolina method after 16 h shaking was 22 per cent less than the quantity extracted after 5 min.