More than 1,000 Spiroplasma isolates have been obtained from horse flies and deer flies (Diptera:Tabanidae) in the United States and Canada. However, the spiroplasma biota of Central America is poorly known. In August of 1995 and 1998, 13 isolates were obtained in 14 attempts from horse flies of a single species, Poeciloderas quadripunctatus, taken in the Costa Rican highlands (1,100–2,000 m). The majority of the “isolates” proved to be mixtures of two or more Spiroplasma species, but after filter cloning, single strains emerged that were designated as representatives of the 13 accessions. Six distinct spiroplasma serogroups were identified from these isolations. Three of the strains are putative new species with no serological relationship to any other Spiroplasma species. A fourth strain is a putative new species that may be distantly related to S. helicoides, a southeastern U.S. species. These four strains are accorded herein status as representatives of new serogroups: strain BARC 4886 (group XXXV); strain BARC 4900 (group XXXVI); strain BARC 4908 (group XXXVII); and GSU5450 (group XXXVIII). A fifth Spiroplasma species was very closely related to S. lineolae, known previously only from the Georgia (U.S.) coast. The sixth was most closely related to subgroup VIII-3, known from Texas and the southeastern U.S. Discovery of six spiroplasma species in only 13 attempted isolations reflects diversity seldom equaled in southeast Georgia, and never elsewhere in the U.S. These results are consistent with a hypothesis that spiroplasma diversity increases from north (Nova Scotia) to south (Georgia and Costa Rica). The discovery of significant affinity between some spiroplasmas of the southeastern U.S. and the Costa Rican highlands was unexpected, but may reflect a climatically complex Pleistocene history.