Phosphorus (P) and metal accumulation in manured agricultural soils and subsequent losses to waterways have been extensively studied; however, the magnitudes and the factors governing their losses during spring snowmelt flooding are less known. We examined the P and metal release from long‐term manured soil to floodwater under simulated snowmelt flooding with recent manure additions. Intact soil columns collected from field plots located in Randolph, Southern Manitoba, 2 weeks after liquid swine manure treatments (surface‐applied, injected, or control with no recent manure addition) were flooded and incubated for 8 weeks at 4 ± 1°C to simulate snowmelt conditions. Floodwater (syringe filtered through 0.45 µm) and soil porewater (extracted using Rhizon‐Mom samplers) samples were periodically extracted and analyzed for dissolved reactive phosphorus (DRP), pH, zinc (Zn), manganese (Mn), iron (Fe), magnesium (Mg), calcium (Ca), and arsenic (As). Mean floodwater DRP concentrations (mg L−1) for manure injected (2.0 ± 0.26), surface‐applied (2.6 ± 0.26), and control (2.2 ± 0.26) treatments did not differ significantly. Despite manure application, DRP loss to floodwater did not significantly increase compared to the control, possibly due to the elevated residual soil P at this site from the long‐term manure use. At the end of simulated flooding, the DRP concentrations increased by 1.5‐fold and 5‐fold in porewater and floodwater, respectively. Metal(loid) concentrations were not affected by manure treatments in general, except for Zn and Mg on certain days. Unlike DRP, where porewater and floodwater concentrations increased with time, metalloid concentration in porewater and floodwater did not show consistent trends with flooding time.