A collaborative multi-institutional program was initiated to streamline the process of microsatellite capture and characterization, development of microsatellites into informative molecular markers, and dissemination of marker information to the cotton research community. A simple and efficient biotin capture method was optimized and used to capture more than 10,000 fragments. Out of 588 fragments sequenced, nearly all contained a microsatellite repeat structure. Several repeat types were represented, including AGA, GA, CA, and ACA. Primers were designed to amplify 307 unique microsatellite loci (305 nuclear and two chloroplast-encoded). One hundred fifty-two microsatellite loci were amplified from G. hirsutum L. cv. TM-1 and Tamcot SP37, and G. barbadense L. cv. Pima 3-79 and Pima S-7. In this comparison, 74 of the primers (~49% of the subset) showed detectable polymorphism. In a comparison of upland G. hirsutum cultivars, ~26% of the primers exhibited intraspecific polymorphism. Polymorphism was widely distributed among the various repeat types and structures (e.g., imperfect and compound repeats). Redundancy with two other previously derived microsatellite marker sets (BNL, CM) was low, implying that the total pool of microsatellites present in the cotton genome is large enough to satisfy the requirements of extensive genome mapping and marker-assisted selection projects.