Cluster analysis in climate studies can define climatic zones and provide valuable information. Validation of the clustering results can be complex, particularly for datasets with no distinguishable boundaries. A proper validity index should consider two main problems: unnecessary cluster merging and unnecessary cluster division to identify precise boundaries. This study demonstrates the weaknesses of three common validity indices (DB, CS, Silhouette) for clustering validation and develops an adjusted DB index (DBR ). The DBR index uses the concepts of spectral clustering and the neighborhood network of data points to control unnecessary merging and division. We use this developed index to validate k-means clustering of precipitation data over the Middle East and North Africa (MENA) to find proper clusters. For comparison purposes, six precipitation datasets, CRU, GPCC, PREC/L, UDEL, CPC daily, and CPC monthly, at comparable spatial resolutions, are examined. Results show that six, six, five, four, five, five clusters delineate the precipitation zones over MENA for the respective datasets. The CRU and GPCC clusters are more consistent with the Köppen-Geiger Classification (KGC) and previous studies. Results indicate that not only the choice of dataset but also the data resolution can affect clustering results. The clustered climate data can be effectively used in adaptation and mitigation planning.