Seed weight and size are important yield components. Thus, selecting for large seeds has been a key objective in crop domestication and breeding. In common bean, seed shape is also important since it influences industrial processing and plays a vital role in determining the choices of consumers and farmers. In this study, we performed genome-wide association studies (GWAS) on a core collection of common bean accessions to dissect the genetic architecture and identify genomic regions associated with seed morphological traits related to weight, size and shape. Phenotypic data was collected by high-throughput image-based approaches, and utilized to test associations with 10,362 SNP markers using multi-locus mixed-models. We searched within genome-associated regions for candidate genes putatively involved in seed phenotypic variation. The collection exhibited high variability for the entire set of seed traits, and the Andean gene pool was found to produce larger, heavier seeds than the Mesoamerican gene pool. Strong pairwise correlations were verified for most seed traits. GWAS identified marker-trait associations accounting for a considerable amount of phenotypic variation in length, width, projected area, perimeter and circularity in four distinct genomic regions. Promising candidate genes were identified, e.g., those encoding an AT-hook motif nuclear-localized protein 8, type 2C protein phosphatases and a protein Mei2-like 4 isoform, known to be associated with seed size and weight regulation. Moreover, the genes that were pinpointed are also good candidates for functional analysis to validate their influence on seed shape and size in common bean and other related crops.