Benefits of long-term No-Till (NT) over Conventional Tillage (CT) include higher infiltration rates and reduced soil erosion, contributing to reduce P export from soils. However, no-till increases P enrichment of the soil surface. This may affect the P loss pattern from soil: although NT can reduce P export related to sediment loss, this practice can promote an increase in dissolved P in runoff. The main purpose of this work was to study how P forms and the P release potential in the surface of a Vertisol are affected in the long-term by soil tillage. An experiment was initiated in 1982 in Carmona, SW Spain (37°24'07" N, 5°35'10" W), involving three tillage treatments with four replications in a randomized complete block design with 180 x 15 m plots. Treatments were: (i) CT involving the use of mouldboard plow to a depth of 50 cm in Autumn and after the use of field cultivator to a depth of 15 cm before sowing; (ii) Minimum Tillage (MT) involving field cultivating to a depth of 15 cm before sowing and partial incorporation of crop residues, and (iii) NT. Crop rotation was wheat-sunflower-pea under rainfed. The fertilizer was only applied to wheat by applying broadcast 240 kg ha^-1 of diammonium phosphate (48 kg P ha^-1) at preplant and 220 kg ha^-1 of urea at sidedress. In CT and MT treatments, the preplant fertilizer was mixed with the soil with field cultivating. In September 2003 a composite soil sample was taken to a depth of 5 cm in each plot. Plant availability index (Olsen P), resin extractable P at 90 days and P forms were studied in each sample. P forms were studied according to the following sequential fractionation scheme: (i) 0.1 M NaOH + 1 M NaCl for 16 h, releases adsorbed P, P precipitated as Fe- and Al-phosphates, and P bound by Fe and Al organic complexes (NaOH-P); (ii) 0.27 M Na citrate + 0.11 M NaHCO₃ (CB) for 16 h, extracts adsorbed P and highly soluble Ca-phosphates, partly precipitated or adsorbed on calcite after the NaOH extraction in calcareous soils (CB-P); (iii) 0.25 M Na citrate at pH 6 and 0.2 M Na citrate pH 6 (C) for 16 h and 8 h, release pedogenic Ca phosphates not dissolved previously (C-P); (iv) 0.2 M Na citrate + 0.05 M ascorbate at pH 6 for 16, releases mostly P occluded in poorly crystalline Fe oxides; (v) 0.27 M Na citrate + 0.11 M NaHCO₃ + 2% Na dithionite for 16 h releases P occluded in crystalline Fe oxides; (vi) 1 M NaOAc buffered at pH 4 for 16, releases residual pedogenic Ca phosphates previously not dissolved by citrate; (vii) 1 M HCl for 1 h, dissolves most lithogenic apatite; (viii) 2 M NaOH at 363 K for 2 h, extracts Organic P; (ix) 0.5 M H₂SO₄ + 0.37 M K₂S₂O₈ for 1 h, extracts Organic P. All the extraction were carried out using 1 g of soil and 40 mL of extractant at 298 K (otherwise indicated). Total P and molybdate reactive P determined in each extract, except step 9. Total P, Olsen P and extractable P by resin at 90 days was significantly higher (approximately double) in NT plots than in CT and MT, thus indicating a higher P release potential in surface soil. A significant increase in the ratio of most labile P fractions (NaOH + CB) to the sum of inorganic P fractions was observed under NT (0.19 NT, 0.1 MT, and 0.09 CT). This relative increase was much higher than that observed for the fraction including low soluble Ca phosphates (C) (0.23 NT, 0.18 MT, and 0.16 CT). Occluded P in Fe oxides to the sum of inorganic P fractions ratio was not significantly increased by NT. Although Organic P was increased in NT plots, the ratio of Organic P to Total P was not significantly increased under NT compared with CT or MT. These results suggest, not only an increase in P content in soil surface, but also an increase in the fraction of this P that could be potentially released to water (runoff water or lost sediments in water reservoirs). Reason for the increased proportion of labile P fractions under NT must be: saturation of sorbent surfaces since fertilizer is accumulating in the surface, and a decreased Ca phosphate precipitation rate due to the effect of the increased organic matter content in NT plots.