A field study evaluating the relative effectiveness of restored and established riparian buffers on shallow subsurface nutrient movement was initiated in 2002 at two NY dairy farms. During the 2004 growing season, we characterized the Morgan extractable phosphorus (P), soil P saturation, and the dissolved reactive P (DRP) concentration in soil water (25 and 50 cm depths) and shallow ground water for 16 paired cropland-buffer plots. Morgan extractable P in buffers averaged 17 kg P ha^-1 across sites and was lower (p<0.01) than cropland (40 kg P ha^-1). Mean soil water DRP at 25 and 50 cm for buffers was lower (p<0.09) than cropland for each sampling. Interestingly, there were no significant differences between 25 and 50 cm soil water DRP in cropland or buffers. Morgan P in corn fields was greater (p<0.01) than hay and buffers, with no differences among hay, grass, grass-willow, and established forested buffer plots. Soil water DRP at 25 cm was strongly correlated with DRP at 50 cm among plots for each sampling, and reasonably approximated soil water DRP at 50 cm. Soil P saturation (oxalate P/ [oxalate Fe + Al]) could be estimated from the Morgan P concentration. Season average soil water DRP at 25 and 50 cm could be estimated by the Morgan P concentration from a curvilinear relationship. Morgan P appeared to be a better indicator of P release to soil water than the soil P saturation or calcium chloride P. Fall ground water DRP concentrations were >10 µg P L^-1 for several corn fields, while average ground water DRP and total dissolved P was lower for buffers. Results indicate that soil P status, soil properties, and landscape variation influence P mobility in shallow subsurface flow and may influence buffer effectiveness over time.