Pollination constant and non-astringent (PCNA) type persimmons, which lose their astringency naturally on the tree, are the most desirable for fresh fruit consumption. The trait of natural astringency-loss in PCNA type is recessive and controlled by a single locus, known as AST locus. Because PCNA type cultivars are of recent origin and exist only in a small number of cultivars, inbreeding depression resulting from repeated crosses among very few cultivars and/or selections is a major challenge facing the current breeding of new PCNA type cultivars. One solution to overcome this inbreeding depression is to include non-PCNA cultivars in the breeding project to extend the genetic pool by utilizing the wide diversity found in non-PCNA cultivars. However, to date, F1 populations resulting from crosses between PCNA and non-PCNA type cultivars/selections have yielded only non-PCNA offspring. To obtain PCNA offspring, F1 progeny needs to be backcrossed to PCNA type cultivars/selections. As persimmon is hexaploid, the obtainable rate of PCNA offspring in a backcross population is only 10-15%. Although we identified AFLP and RFLP markers tightly linked to the trait of natural astringency-loss in persimmon, these markers could not distinguish some non-PCNA progenies in other breeding populations. Previously, we demonstrated that RFLP markers derived from fosmid clones of D. lotus, diploid relatives of D. kaki, could be used to distinguish between PCNA and non-PCNA types of D. kaki, and established the possibility of extending this approach to PCR based markers. In this study, we describe the development of PCR-based markers derived from fosmid clones of D. kaki and D. lotus, and confirm their effectiveness across diverse progenies to select PCNA type.