Weather events where a dry period is followed by a heavy rainfall event appear to affect phosphorus (P) exports through preferential flow pathways from forest soils. Export rates also depend on the P stocks. To explore this, we installed zero-tension lysimeters in three trenches at two sites with contrasting soil P stocks. Lysimeters were installed in three different depths (topsoil, subsoil and deep subsoil) to explore P depth transport. We covered the forest floor above the lysimeters with tarpaulins to simulate a dry period and afterward artificially irrigated the area. This experiment was repeated three times at each site. Lysimeter samples were analyzed for concentrations of total P, organic and inorganic dissolved P and particle bound P (>0.45 μm). Loads of P and flow rates were calculated. Results reveal clear differences between sites, individual events and soil depths. At both sites, concentrations and loads of P in the topsoil lysimeters were higher than those in the subsoil. This difference was most evident at the low P site and underlines its efficiency of recycling nutrients. Dissolved inorganic P showed marked peaks in the topsoil lysimeters, whereby in the subsoil, particle-bound P peaks were partly noticeable at both sites. Depth transport of P into the subsoil depended on initial soil moisture, texture and the spatial distribution of flow pathways. Further, we observed large heterogeneity within a single site, dependent on profile-specific characteristics of the distribution of P, flow pathways and microbial biomass. We conclude that under certain conditions, there is a depth transport of P into the subsoil and therefore a potential of P exports, especially for particle-bound P. Small-scale heterogeneity hampers the clear identification of influences and illustrates the need for further research regarding soil heterogeneity.