Random amplified polymorphic DNA, simple sequence repeat, and inter-simple sequence repeat markers were used to estimate the genetic relations among 65 pea varieties (Pisum sativum L.) and 21 accessions from wild Pisum subspecies (subsp.) abyssinicum, asiaticum, elatius, transcaucasicum, and var. arvense. Fifty-one of these varieties are currently available for growers in western Canada. Nei and Li's genetic similarity (GS) estimates calculated using the marker data showed that pair-wise comparison values among the 65 varieties ranged from 0.34 to 1.00. GS analysis on varieties grouped according to their originating breeding programs demonstrated that different levels of diversity were maintained at different breeding programs. Unweighted pair-group method arithmetic average cluster analysis and principal coordinate analysis on the marker-based GS grouped the cultivated varieties separately from the wild accessions. The majority of the food and feed varieties were grouped separately from the silage and specialty varieties, regardless of the originating breeding programs. The analysis also revealed some genetically distinct varieties such as Croma, CDC Handel, 1096M-8, and CDC Acer. The relations among the cultivated varieties, as revealed by molecular-marker-based GS, were not significantly correlated with those based on the agronomic characters, suggesting that the 2 systems give different estimates of genetic relations among the varieties. However, on a smaller scale, a consistent subcluster of genotypes was identified on the basis of agronomic characters and their marker-based GS. Furthermore, a number of variety-specific markers were identified in the current study, which could be useful for variety identification. Breeding strategies to maintain or enhance the genetic diversity of future varieties are proposed.