Soil surface roughness has a strong influence on runoff and erosion, affecting surface storage capacity, water flow routing and velocity, and modifying runoff rates. It also reduces soil detachment by raindrops and the shear effect of water flow on the soil surface. In arid and semiarid ecosystems, biological soil crusts (BSCs) commonly appear in clearings between plants. Depending on the dominant component in the BSC community, the roughness of the soil surface may vary considerably, changing the hydrologic and erosive response of the soil. The aim of this study is to determine the effect of physical and biological crusts on soil surface roughness and their influence on runoff and erosion. For this purpose, we set up open plots containing different BSC types and treatments and recorded runoff and total erosion from all the events for 18 months. Micro-topographic indexes were calculated from high-resolution digital surface models of the plots built from terrestrial laser scanner height data. After comparing different spatial resolutions and indexes, we concluded that only the local Random Roughness index in a 40-mm moving window provided a precise estimation of the roughness induced by BSCs, though it did not have a direct effect on runoff response. The best relationship between microtopography and runoff on biologically crusted soils was found for surface storage capacity, which appears as a powerful predictor of the runoff coefficient on long temporal scales. Sediment yield was not well predicted by any of the micro-topographic indexes studied. The only index that was significantly related to sediment yield was the local Random Roughness in a 40 mm moving window, but even this explained only a third of the erosion variance.