Fourteen SSR markers previously developed for cultivated olive, Olea europaea ssp. europaea var. europaea , were used to genotype over 300 accessions of cultivated olive from five collections of 19^th-20^th century plantings in California and the University of Arizona, in order to examine genetic diversity, structure, and differentiation within and among collections. Genotypes of cultivars from the National Clonal Germplasm Repository for Fruit & Nut Crops (NCGR) collection were compared with the other collections to potentially establish cultivar identity. Up to 90% of samples per non-NCGR collection were found to be synonymous with named, primarily Spanish, cultivars, in agreement with the history of early olive importation and development in California. The number of alleles per locus across all collections ranged from one at UD099-019 in the Santa Barbara collection to 16 at ssrOeUA-DCA11 in the NCGR collection, with a mean number of alleles per locus per collection ranging from 2.9 for Santa Cruz Island and 9.9 for the NCGR collection. Several different loci in the non-NCGR collections were either completely homozygous (H_o = 0.00) or completely heterozygous (H_o = 1.00), while the mean observed heterozygosity per collection ranged from 0.598 for the University of Arizona to 0.650 for Lotusland Gardens. Contingency chi-squared analysis showed significant differences in allele composition and frequencies across all loci among the collections. Many alleles found at low frequencies within the NCGR collection yet missing entirely within the other collections suggest that the NCGR collection contains greater diversity, and cultivars that were not widely distributed throughout California in the 19^th and early 20^th centuries. The unique genotypes found within the non-NCGR collections were distributed throughout the UPGMA phenogram and generally most closely related to one or more named cultivars, with a few exceptions. Projection of accessions onto the first two or three principle components (26% of variation) generally agreed with the results of the UPGMA phenogram. The results suggest that historical olive collections in California and Arizona, while containing small numbers of unique genotypes, may still contain significant genetic diversity, including genotypes and alleles no longer present in the modern California olive industry or official germplasm collection.