The low-chill southern highbush blueberries now being grown in Florida were developed by crossing northern highbush cultivars, which had large fruit, high fruit quality, and early ripening but were poorly adapted in Florida, with Florida native blueberry species, which were well-adapted but were late-ripening and had small fruit. In formulating the foundation populations for recurrent selection during the 1950s, most of the parentage was northern highbush, and less was Florida native. This paper presents the hypothesis that faster progress could have been made if a higher percentage of the original gene pool had been derived from the adapted natives. The hypothesis is based on the observation that good environmental adaptation depends on many independent components, most of which are hard to evaluate accurately in small plots in one year, whereas large fruit, high fruit quality, and early ripening are easier to evaluate in single-plant plots in a single year. Them ability to evaluate young plants in small plots allows rapid cycles of recurrent selection during which the selected characters can quickly be improved. Florida native blueberry species that could have been used more in breeding include diploid Vaccinium darrowi Camp, diploid highbush blueberry (V corymbosum L.) from the Florida peninsula between Ocala and Lake Okeechobee, and tetraploid highbush blueberry (V corymbosum L.) from northeast Florida and southeast Georgia.