SummaryWe wished to determine whether soil‐test P was affected by storing air‐dry soil samples at room temperature. The soil samples had been collected from field experiments and air‐dried (<40°C) before measuring soil‐test P (bicarbonate‐extractable P). The samples were from field plots that had been treated with different applications of fertilizer P (superphosphate, rock phosphate) one or more years previously. Soil‐test P was measured on two different sub samples of the same sample: either A, in the year the sample was collected; or B, after the sample had been stored at fluctuating room temperatures either from 2 to 8 years (four field experiments) or 17 years (59 field experiments). The room temperature ranged from 10 to 30°C, and averaged 17°C. The aim was to test whether soil‐test P was systematically and consistently different between sub samples A and B.Differences between A and B were mostly small, and there were no consistent or systematic differences. For the Colwell soil test, applied to a range of south‐western Australian soils, possible decreases in soil‐test P due to continued reaction with the soil could not be detected using bicarbonate‐extractable soil P, and storage of air‐dry samples at room temperature did not significantly affect soil‐test P measured up to 17 years later. We conclude that, provided fertilizer P has had time to react with soil in the field, no further changes in Colwell soil‐test P occur during air‐dry storage for up t