Suspended Sediment (SS) and Phosphorus (P) are two important pollutants transported in runoff from agricultural land that can cause deterioration in the quality and biodiversity of surface waters. To maintain water quality for a range of users at the local level, national regulators and catchment stakeholders are increasingly seeking simple methods to help identify where diffuse pollutants such as SS and P originate within a catchment. Since soil erodibility is a function of the properties of the soil, a number of physically-based soil tests related to soil aggregate stability, dispersibility and shear strength have been developed to help characterise vulnerability to soil erosion under different site conditions. Similarly, routine soil P tests have historically been used to help distinguish over-fertilised soils and more recently the risk of soluble P release to runoff. However, soil tests that determine the risk of P mobilisation in particulate form are lacking, despite the fact that the largest losses of P in land runoff are usually associated with eroding soil particles. Soil erodibility tests based on dispersion therefore have the potential to provide information on the P signal associated with different sizes of eroding particles in surface and sub-surface runoff from agricultural land, and it is appropriate to consider SS and P loss risk in combination. This study investigated the feasability of a combined environmental soil test that could distinguish the intrinsic risk of both SS and P mobilisation from different soils, and provide data that could be incorporated into decision support tools for catchment management. Twenty-six representative bulked samples (0-10 cm) of topsoil collected from selected experimental field sites across Europe (hereafter called the 'benchmark' soils) were slowly air-dried, gradually broken up by hand and gently sieved to produce aggregates <5 mm in size. These samples were analysed for their physical, mineralogical and chemical properties. The potential mobilisation of soil particles and P forms in runoff from the benchmark soils was measured under controlled conditions indoors using simulated rainfall from a height of 9 m. Each soil was uniformly packed into a tray (0.5 m long, 0.25m wide and 0.085 m deep) with a 5° slope, prewetted for 24 hours and then subjected to two successive rainfall events (60 mm hr^-1 for 30 minutes). The overland flow was collected and subsamples taken for determination of SS, total P (TP) and dissolved (<0.45 µm) P fractions. A range of aggregate stability and soil dispersion tests were developed in the laboratory to help explain differences in the amounts of SS and P forms collected under indoor rainfall simulation. Aggregate stability tests included a wet sieving index, single drop, percolation stability and soil consistency. Three alternative methods of soil dispersion were compared in the laboratory: gentle dispersion in water, chemical dispersion in sodium chloride and ultrasonic dispersion. Comparison of the amounts of SS and P fractions obtained by laboratory tests with those obtained in runoff under rainfall simulation indicated that the gentle water dispersion method gave the most reproducible and predictable results. Linear regression coefficients (r²) were 0.67 for SS, 0.79 for TP and 0.81 for DP. The combinations of shaking time, settling time and soil:solution ratio which gave the best correlations with sediment and P loads under indoor rainfall simulation were 1 min, 4min 40s (corresponding to <20 µm fraction) and 1:50 ratio, respectively. Both the indoor runoff and laboratory dispersion test results indicated that the risk of sediment mobilisation was most closely and negatively related to soil organic matter (r²=0.4), and that DP (but not PP) concentrations could also be predicted from Olsen-extractable and water-extractable P concentrations in the soil. The potential applications of the environmental soil test developed are discussed.