: In recent years, many pastoral farmers in New Zealand reduced or withheld phosphorus (P) fertilizer application to their pasture in response to rising fertilizer costs and falling prices in farm produce. When P fertilizer application is withheld, pasture and animal production relies on the P reserves accumulated in the soil from the previous P fertilizer applications. A field‐plot experiment superimposed on irrigated pastures that had received long‐term annual superphosphate (SP) fertilizer applications for 25 years was conducted to examine the forms of P accumulated from the previous SP applications contributing to plant‐available P. The long‐term SP applications were withheld, and the soil was subjected to treatments of no fertilizer (nil), P only, sulphur (S) only, or both P and S, each with P and S at two rates in a split‐plot design with four replicates that ran for 6 years. Soil samples (0–75 mm deep) collected from the trial each year were subjected to a sequential P fractionation scheme. Results showed that most of the P forms extracted by the different extractants showed no significant differences between treatments. Only the inorganic P (Pi) extracted by sodium bicarbonate (NaHCO3), sodium hydroxide (NaOH) I, and hydrochloric acid (HCl) showed significant differences, and these were related to plant‐available P. However, it was not possible to isolate an individual P form as the dominant P form accumulated from previous SP applications as the plant‐available P form.