Abstract. In the context of sustainable agricultural management, drought monitoring plays a crucial role assessing the vulnerability of agriculture to drought occurrence. Drought events are very frequent in the Iberian Peninsula (and in Portugal in particular) and an increase of frequency of these extreme events are expected in a very near future. Therefore, the quantitative assessment of the natural ecosystems vulnerability to drought is still very challenging, mainly due to the difficulties of having a common definition of vulnerability. Consequently, several methods have been proposed to assess agricultural vulnerability. In this work, a principal component analysis (PCA) was performed based on the components which characterize the Exposure, Sensitivity and Adaptative Capacity of the agricultural system to drought events with the aim of generating maps of vulnerability of agriculture to drought in Portugal. Several datasets were used to describe these components, namely drought indicators, vegetation indexes and soil characterization variables. A comparison between the PCA-based method and a subjective non-automatic categorical method using the same indicators was performed. Results show that both methods identify Minho and Alentejo as regions of low and extreme vulnerability, respectively. The results are very similar between the two methods, with small differences on certain vulnerability class. However, the PCA method has some advantages over the categorical method, namely the ability to identify the sign of the indicators, not having to use the indicator-component subjective relationship, nor calculating weights. Furthermore, the PCA method is fully statistical and presents results according to a prior knowledge of the region and the data used.