Stabilization of soil surface condition by conservation tillage and/or by using soil amendments might be effective in decreasing runoff and erosion as well as on nutrient loss from most of cultivated land that has received long-term fertilizer or manures. For many of US field soils, the breakdown of surface aggregates results in pore collapse, changes in macro- and micro aggregate stability (slaking, mechanical and/or chemical dispersion), causing runoff and soil loss from the soil surface. We assumed that by preventing slaking and/or dispersion, it is possible to control the mechanism of phosphorous (P) entrainment. We studied the effect of soil surface condition (degradation or disturbance level) and surface application of soil amendments (gypsum and polyacrylamide, PAM) on phosphorous (soluble, SP or attached TP) loss from a silty loam (fine-loamy, mixed, mesic Typic Hapludalf) soil. Erosion flumes (100 long x 20 wide x 10 deep, cm) were packed with <4 mm diameters aggregates (mixed with a known amount of P, fertilizer) and then wetted fast or slowly with distilled water using a peristaltic pump. The treatments were: control (aggregate breakdown with fast wetting, representing greater degradation under intensive tillage); slow wetting (lower degradation, representing no till); gypsum (2 t ha^-1); and PAM (20 kg ha^-1 + gypsum (2 t ha^-1), i.e. minimizing dispersion and/or slaking and disintegration of surface aggregates under raindrop impact. Amendments were applied on the surface of fast wetted soil prior rainstorm with high energy and intensity (60 mm h^-1) and 1h duration. Adding the amendments to the fast wetted soil surface and/or keeping soil surface stable by slow wetting: (i) increased surface aggregate stability and drainable porosity with the pattern of control < gypsum < slow wetting < PAM + gypsum; and decreased runoff and soil loss with the same pattern, suggesting that runoff or soil loss was a function of aggregate stability; (ii) resulted in a different shaped curve the P loading dynamics; and changed the SP and TP concentrations differently with from the aggregate stability pattern. This indicated that P loading also depends on soil aggregate breakdown mechanism. Similar to soil loss, PAM + gypsum treatment was more effective in reducing TP and SP concentrations (> 1.5-3 time), whereas using of gypsum alone was more capable on lowering SP concentration (> 2 time). Slow wetting increased SP and slightly affected the TP concentration. Total P loss was greatest for control and minimum at PAM + gypsum or gypsum treatments (< 2-4 times). Generally increase in P fertilizer rate increased the SP loss in runoff. Slowly wetted aggregates, being less detachable, transferred more SP from soil solution to runoff or infiltration water before a completed seal developed, whereas, for control treatments the opposite was true for the TP. The PAM + gypsum treatment stabilized macro-and micro aggregate stability and drainable porosity (i.e. by the decreasing detachment and transport of particles, flocculating dispersed clay particles) of the soil surface and decreased both the concentration of SP and TP. This was also due to the conversion of SP to the insoluble Ca-bound pools by gypsum dissolution. The use of either PAM + gypsum or gypsum alone in addition to simulated (tilled) soil surface showed considerable efficacy in reducing the feasible pollution of surface water quality. The effect of the treatments on surface condition and P loss in runoff is discussed in this paper.