The objectives of this study were to examine the variability in mobilization and transport of primary particles and associated total phosphorus (TP) in sediments eroded by overland flow from 13 European arable soils and to consider the empirical support for more process-based alternatives to modeling phosphorus (P) transfers. The 13 soils were subjected to simulated rainfall in laboratory experiments. Rainfall was applied to a soil flume (0.5 × 0.25 m) for 30 min at intensity of 60 mm h1, and all overland flow generated during this period was collected. Two simulations were performed 5 d apart. The soils generated a wide range of overland flow (13.3–26.9 mm) and sediment (1.1–16.9 g). The sediments from the experiments were enriched with medium silt particles (6–20 µm). Except for one soil in the second simulation, all of the study soils produced overland flow sediments (OFS) that were enriched with P (TP 976–3884 mg kg1, P enrichment ratio 0.92–4.42). Sediment TP was positively correlated (P < 0.05) with the sediment clay contents for both simulation events, and sediment total P enrichment was negatively correlated with soil TP. Attempts to reconstruct the TP content of sediments from TP analysis of the soil particle size fractions yielded significant (P < 0.05) relationships between reconstructed and measured TP values. However, the reconstructed TP values were consistently lower than the measured sediment TP contents, indicating that further improvements to measurement techniques are required if physically based modeling of P transport in overland flow is to be successful at small scales