Urban composts represent a significant reservoir of nutrients, particularly phosphorus (P). Their application in agriculture requires more information on their impact on the dynamics of soil P. In this study, the long-term effects of repeated applications of urban composts and manure on the dynamics of organic P (orgP-SW) and inorganic P (inorgP-SW) stocks using the Saunders and Williams’ (SW) method were analyzed in the soil of the Qualiagro field experiment (4 blocks), grown in a corn / wheat succession. Five treatments were studied: a control without P (CONTROL), a cattle manure (MANURE) and 3 urban composts: compost green waste + sludge (DVB), a biowaste (BIO) and residual household waste (OMR). They were added to soils on the basis of 4 t C / ha every two years over a period of 15 years. We determined the orgP-SW content in the applied products and the plough layer by difference between P extracted by sulfuric acid (0.1M) in a calcined sample (totP-SW) and not calcined (inorgP-SW) samples. All products mainly contained inorgP-SW (81%) because orgP-SW averaged 19% for the 3 urban composts and manure. The average orgP-SW soil content in 2013 was 31% of Ptot-SW and did not differ among treatments. After 15 years, cumulated balance of added minus exported P varied between - 351 (± 12) kg P ha-1 in CONTROL and +1292 (± 21) kg P ha-1 in the DVB treatment. Despite these differences, the stock of total organic P was invariant across most treatments but in the manure treatment where it was higher. Variations in inorgP-SW and total P stocks reflected differences in P balances. The stability of soil orgP-SW in CONTROL indicated that soil organic P contributed little or not at all to crop phosphate nutrition.