PM 10 and PM 1 samples were collected at an urban site in southeastern Spain during 2006e2010. The chemical composition of all samples has been determined and analyzed by Positive Matrix Factorization (PMF) technique for fine and coarse source identification. The PMF results have been analyzed for working and non-working days in order to evaluate the change in PM sources contribution and possible future abatement strategies. A decreasing trend in PM 10 levels and in its constituents has been observed, being partly associated to a reduction in anthropogenic activities due to the economic crisis. The use of fine and coarse PM in the PMF analysis allowed us for the identification of additional sources that could not be identified using only one size fraction. The mineral dust source was identified in both fractions and comprised 36 and 22% of the total mass in the coarse and fine fractions, respectively. This high contribution of the mineral source to the fine fraction may be ascribed to contamination of the source profile. The regional re-circulation source was traced by secondary sulfate, V and Ni. It was the most important source concerning PM 1 mass concentration (41% of the total mass in this fraction). Although V and Ni are commonly associated to fuel oil combustion the seasonality of this source with higher con-centrations in summer compared with winter suggest that the most important part of this source can be ascribed to regional pollution episodes. A traffic exhaust source was identified but only in the fine fraction, comprising 29% of the fine mass. The celestite mines source associated with nearby open-pit mines was typified by strontium, sulfate and mineral matter. PM 10-1 levels were higher in working days, whereas PM 1 levels remained fairly constant throughout the whole week. As a conclusion, traffic seems to be the main source to target in Granada.