The assessment of impacts of an altered nutrient availability, e.g. as caused by consistently high atmospheric nitrogen (N) deposition, on ecosystem phosphorus (P) nutrition requires understanding of P fluxes. However, the P translocation in forest soils is not well understood and soil P fluxes based on actual measurements are rarely available. Therefore, the aims of this study were to (1) examine the effects of experimental N, P, and P+N additions on P fluxes via preferential flow as dominant transport pathway (PFPs) for P transport in forest soils; and (2) determine whether these effects varied with sites of contrasting P status (loamy high P/sandy low P). During artificial rainfall experiments, we quantified the P fluxes in three soil depths and statistically analyzed effects by application of linear mixed effects modeling. Our results show that the magnitude of P fluxes is highly variable: In some cases, water and consequently P has not reached the collection depth. By contrast, in soils with a well-developed connection of PFPs throughout the profile fluxes up to 4.5 mg P m⁻² per experiment (within 8 h, no P addition) were observed. The results furthermore support the assumption that the contrasting P nutrition strategies strongly affected P fluxes, while also the response to N and P addition markedly differed between the sites. As a consequence, the main factors determining P translocation in forest soils under altered nutrient availability are the spatio-temporal patterns of PFPs through soil columns in combination with the P nutrition strategy of the ecosystem.