Cluster analysis is a type of exploratory analysis that is used for classifying unknown individuals into groups of members that share certain similarity. Grouping breeding materials into different clusters based on their performances under a given condition allows plant breeders to select breeding lines more efficiently. This study aimed to group breeding materials based on their performance under water stress conditions using cluster analysis. The experiment was conducted in RCB design with 3 blocks. Fifty maize inbred lines were grown and exposed to water stress conditions. Twenty-four phenotypic traits were collected and some of them were subjected to cluster analysis. The Partitioning Around Medoid algorithm was used to cluster 50 inbred lines from chosen phenotypes. Cluster validation was then carried out in a subsequent experiment by testing the statistically significant differences between chosen inbred lines and tolerant and susceptible clusters. According to the analysis, 4 clusters with different numbers of inbred lines were obtained. Lines in cluster no. 3 showed the most tolerance compared with the other clusters. In contrast, individuals in cluster no. 1 were the most susceptible. The results of cluster validation in another year also supported this cluster result. Therefore, we concluded that the clustering method was an efficient way to differentiate tolerant and susceptible inbred lines. Furthermore, the results suggested that plant height, tassel size, spikelet density, days to silking, and anthesis-silking interval were secondary traits that could be used in selection for drought tolerance in maize.