It is now well established that mating can occur between diploid a andα cells ofCandida albicans. There is, however, controversy over when, and with what efficiency, nuclear fusion follows cell fusion to create stable tetraploid a/α cells. In this study, we have analysed the mating process betweenC. albicansstrains using both cytological and genetic approaches. Using strains derived from SC5314, we used a number of techniques, including time-lapse microscopy, to demonstrate that efficient nuclear fusion occurs in the zygote before formation of the first daughter cell. Consistent with these observations, zygotes micromanipulated from mating mixes gave rise to mononuclear tetraploid cells, even when no selection for successful mating was applied to them. Mating between different clinical isolates ofC. albicansrevealed that while all isolates could undergo nuclear fusion, the efficiency of nuclear fusion varied in different crosses. We also show that nuclear fusion inC. albicansrequires the Kar3 microtubule motor protein. Deletion of theCaKAR3gene from both mating partners had little or no effect on zygote formation but reduced the formation of stable tetraploids more than 600-fold, as determined by quantitative mating assays. These findings demonstrate that nuclear fusion is an active process that can occur inC. albicansat high frequency to produce stable, mononucleate mating products.