Abstract: This study aims to clarify the influence of forests, as well as other prevalent land cover types, on extreme hydrological events through a land cover gradient design. We selected 10 catchments within a gradient of forest land cover, in which there were 15 years of simultaneous daily hydrological and meteorological data, and an additional forest descriptor, forest maturity. The study was developed in a heterogeneous region in the Cantabrian Mountains (NW Spain). This area includes different vegetation types and has a long history of human disturbance and land use change that has produced a gradient in forest cover. This study focuses on regular hydrological extremes: regular floods and low flow events. Specific objectives were to observe the relationship between land cover and extreme hydrological events, once the variance explained by precipitation was removed, and compare the effectiveness of forest coverage and maturity to predict them. Partial correlations and ordinary least square regressions were developed using hydrological indices, obtained from flow records, and hydrological parameters calculated through modelling, using the Identification of unit Hydrographs And Component flows from Rainfall, Evaporation and Streamflow data (IHACRES) software and hydrometeorological data. Land cover characteristics were better able to predict floods than low flows. Forests were associated with less extreme flow events (lower intensity and frequency of floods and greater base flows), whereas shrub formations did the opposite. These results were more evident using forest maturity than using forest coverage. This study indicates that hydrological modelling may benefit in the future from considering not only the coverage of different land cover types but also the conservation status of the different vegetation formations.