A broad-scale analysis of the structure and understory composition of Pyrenean mountain pine (Pinus uncinata Ram.) stands was performed using data from the Spanish National Forest Inventory. Twelve structure-based forest typologies were defined from variables related to tree size, stand density, vertical structure and standing deadwood, using cluster analysis techniques. These typologies were adequately classified (accuracy >75 per cent) by a dichotomous key obtained from classification and regression trees. Multiple regression models were then used to analyse relationships between the main stand structural variables and a set of climatic and physiographic factors. The models showed significant correlations between winter temperature, slope and continentality (among other variables) and the current structure of mountain pine stands. The relationships between the understory composition of mountain pine forests and different environmental and structural overstory factors were found to be driven by an elevation-pH gradient and a stand density-soil stoniness gradient. The results of this study can be directly used for forest planning at different scales and could help forest managers to establish strategies designed to facilitate a given habitat for species of conservation interest.