This study uses principal component analysis (PCA) and clustering (CL) techniques to characterize the daily surface synoptic circulation patterns over the region 30°N-60°N by 30°W-15°E, for the period 1960-2001 using the NCEP/NCAR Reanalysis Project data. The methodology used, previously proposed by Esteban et al. (2005) to identify heavy snowfall days over Andorra (Pyrenees), involves a pre-processing approach consisting of a spatial standardization of the data used for the rotated PCA on S-mode analysis and correlation matrix, and Varimax rotation of the retained components. For the clustering process, an alternative approximation based on the PCA results (positive and negative phases) to decide the centroids and the number of groups for the K-means clustering is shown, followed by the rejection of the iterations for this non-hierarchical algorithm. Twenty sea level pressure (SLP) circulation patterns were obtained, with all days classified. The composite maps for SLP and 500 hPa geopotential height (GPH), the monthly distribution, and the long-term variability for each of the circulation patterns is obtained. The results are consistent with the subjective knowledge of the daily atmospheric circulation over the area treated, and seem to be an easy and accessible way for the synoptic studies of short-lived (from weekly to daily) phenomena. In addition, this methodology appears to be applicable to all climatic areas of the world, as it characterizes well enough the complex circulation variability at mid-latitudes, including the surface low-pressure gradient patterns normally located over the Mediterranean in summer.