The temporal occurrence of slope movements, their contribution to relief evolution and human influence on those processes in two study areas of northern Spain are presented. The work is based on temporal analyses covering 100 ka in one study area and 43 years in the other. Temporal analysis has been the basis for quantitatively assessing the magnitude of human influence and developing landslide susceptibility and hazard models with known, independently-tested prediction capability. The results obtained in one study area show a relationship between landsliding periods and increasing precipitation during upper Pleistocene and Holocene. Significant increases of landslide frequency and mobilisation rate were also found around 5500 and 200 years ago. Those moments coincide with two periods of intensified human presence and activities: Neolithic and industrial revolutions. The increase observed represents about one order of magnitude from pre-Neolithic to present. A similar increase has been found between 1954 and 1997 in the other study area. The latter increase shows no relationship with changes in climate parameters or seismic activity. A fairly good correlation has been found between landslide frequency and socioeconomic indicators of human activity. Sedimentation rates in two neighbouring estuaries were determined and significant increases, particularly in the second part of last century, were also found. Evidence obtained suggest that the increases observed in the frequency of slope instability events (and therefore hazard), denudation and sedimentation rates are due to a greater extent to indirect geomorphologic changes caused by human action rather than climate change. Detailed analysis of landslide frequency during that 43-year period has also been used to produce and validate landslide susceptibility models and obtain landslide hazard maps with known prediction accuracy. Validation tests were carried out comparing susceptibility maps based on landslides that have occurred in the past with the ones occurring in several, later periods. Prediction capability of models could thus be determined. Validation of susceptibility maps using data derived from the temporal analysis of landslides was used to obtain hazard maps expressed as probability of new ruptures in a given time period, on the basis of past trends in terrain behaviour. Possible impacts of a growing influence of human activities on the terrain have been incorporated into the assessment of future hazards by considering different scenarios.