This study investigated the effects of different clonal strains of Symbiodinium sp. (zooxanthellae) on clam growth and survival, while monitoring the persistence of the induced symbioses in outdoor tanks and in the field using allozyme and random amplified polymorphic DNA (RAPD) analyses. Aposymbiotic clam larvae that were inoculated with homologous zooxanthellae (cultured or freshly isolated from the same host species) or heterologous zooxanthellae (cultured from different host species) had significantly-different survival rates at harvest (3 months post-spawning) with small growth differences. The improved survival rates in juvenile Hippopus hippopus (heterologous infection) and Tridacna gigas (homologous infection) were maintained through 3 months onshore and 3 months offshore. However, isozyme and RAPD analyses of re-isolated zooxanthellae after 3, 6, and 9 months revealed a high genetic diversity of symbionts (ca. 99% variation in 200 re-isolates) from individual hosts, within and between treatments. Furthermore, the genetic patterns of the re-isolated algae following clonal culture were different from those of the introduced clones, which, in contrast, retained their unique genetic patterns over many culture generations in the laboratory. These results demonstrate that the subsequently-established symbiont populations in juvenile clams were not clonal. The allozyme and RAPD techniques detected individual genetic differences in clam symbionts, but not differences between algal taxa. The presence of significant survival trends suggests possible differences between subsequently-established dominant symbiont taxa, which were uncultureable or undetectable using these genetic markers. The implications of this symbiont diversity in giant clams are discussed.