Collecting germplasm from even a fraction of the potential geographic range of wild potato species requires substantial time, money, and effort, so efficiency could be increased if spots particularly rich in unique alleles could be predicted and prioritized. A previous experiment that used AFLP markers to compare "remote" versus "easy" collection sites for Solanum stoloniferum, (previously S. fendleri) within three mountain ranges identified the Santa Catalina Mountains (CAT) of SE Arizona as making a particularly large contribution of unique alleles, despite existing CAT collections being few and all close to roads. This situation motivated a collecting expedition in September 2009 to more thoroughly collect CAT. That expedition resulted in samples of 16 populations, most from new sites never previously described or collected. An analysis was done with 871 AFLP loci, comparing populations from the same three mountain ranges and including the 2009 CAT collections. Results confirmed the prediction of unique allele density of certain location categories within mountain ranges. The new CAT collections captured three times as many unique alleles as contained in the previous collections from that location, and 24 of these were new (unique) among all mountain ranges. One particular new collection, PI 658180, accounted for 46% of the new unique alleles collected in CAT. This study demonstrates the power of DNA markers to empirically identify locations with genetic richness, guiding the most efficient allocation of resources for collecting, preservation, and evaluation of germplasm. © 2011 Potato Association of America.