Dissolved N and P transfer to runoff water may increase with surface applications and shallow soil incorporation of animal manure. Information is needed regarding water-extractable nutrient release during manure decomposition to quantify that potential transfer to runoff in permanent pastures and conservation tillage systems. Release of net mineralizable C (MIN_C), net mineralizable N (MIN_N), and dissolved reactive P (DRP) was determined in stockpiled and composted cattle (Bos taurus) manure and manure-amended soils at 4, 20, and 35degreesC for 322 d at about 60% water-filled pore space. Flushes of CO2-C exceeding 100 mg kg(-1)d(-1), inorganic N, and DRP were released rapidly from both manures when incubated alone or as soil amendments. Dissolved P release varied inversely with sorption capacity and degree of P saturation in an Aridic Paleustalf and Torrertic Paleustoll. Net mineralizable C, MIN_N, and DRP flux densities were lognormally distributed during the 322-d incubation. Results from the lognormal modeling approach suggest that incubations needed to be performed only for as long as needed to attain the 50% maximal flux density beyond the maximum to predict MIN_C, MIN_N, and DRP release flux density distributions. Significant nonlinear relationships exist between ln(cumulative CO2-C) and inorganic N or DRP and have an inflexion point between 14 and 20 d. The nonlinearity of the C-to-N and C-to-DRP relationships indicates multiple substrate pools and supports the use of lognormal distributions to describe MIN-C, MIN_N, and DRP release from manures and manure-amended soils and to shorten laborious incubations.