ABSTRACT Many of Britain's reclaimed coal-lands are damaged by erosion. Much accelerated runoff and erosion is caused by soil compaction that reduces the land's capacity to absorb rainwater. This study compares conditions in two types of coal-spoil, deep-mine colliery discard and surface-coal-mine overburden, and in four contexts: freshly laid spoil, old unvegetated minespoil, minespoil partly veneered by grass tufts, and ‘reclaimed’ spoils, which are defended by a dense grass turf and (ca. 100 mm) deep artificial topsoil. Bulk density testing reveals a characteristic pattern. The soil is of moderate density (< 1.3 g/cm3) in the grass root zone. However, soil densities rise to 1.8 g/cm3 and more between 30 and 50-cm depth. Particle size and stability tests confirm that minespoils contain a large proportion of water unstable primary particles, mainly mudstones and shales. These break down on wetting, flood the soil with fine particles, and help raise the soil density to levels where water infiltration and root penetration is impeded. Rainfall simulation tests on surface coalmine spoils near Blaenavon, Wales, find that poorly vegetated plots (50-90% cover) convert less rainfall to runoff, and yield a fourth as much sediment as unvegetated plots. Tests of deep coalmine spoils at the Earth Centre Reclamation Project, Doncaster, England, found that newly laid, unvegetated, recently washed and hence stabilised, coal spoils converted only 20% of rainfall to runoff. Older comparted, spoils converted 25-35% of the rainfall to runoff when unvegetated but less than 10% when mantled by thin grass turf. Soil losses for the same rain event were more than four times greater on the unvegetated sites. Finally, on both sites, plots mantled with a deep well-managed layer of applied topsoil sites absorbed most of the rainwater applied in tests. This was subsequently detected running in macropores and fissures below the applied topsoil and in the upper layers of the relatively impermeable mine-spoil subsoil.