It has been suggested that the capacity of fusion with both kin and genetically unrelated conspecifics to form chimeras (i.e. individuals wlth a mixture of genetically different cells) is evolutionarily retalned In several phyla because the resulting organism obtains some selective advantages over non-chimeric conspecifics. Many demosponges are known to have fusible larvae that form young chimeric sponges, but the ecological and evolutionary significance of this phenomenon has seldom been investigated. It is reasoned here that if chimeras have a selective advantage, their formation will be expected to be favored by mechanisms promoting larval encounters at settlement. By using sibling larvae of the demosponge Tedania ignis in the laboratory, I tested the hypothesis that larvae show a natural tendency to aggregate and form chimeras In a 50 d field-transplantation experiment using chimenc sponges obtained from the fusion of 2 sibling larvae, 1 also tested the hypothes~s that size and survival are increased in chimeric ind~viduals compared to non-chimeric ones and investigated the permanent versus transitory character of this chimensn~. It was found that larvae did not show any significant tendency to settle spontaneously near siblings, and no fusion between s~blings took place. When pairs of larvae were forced to settle in contact, fusion was, however, the outcome in all cases. It was also found that, although chimeric sponges were stable and about twice as large as non-chimeric sponges, they did not show increased survival. These results disprove the common argument that the capacity of fusion with conspecifics is evolutionarily retained in many sponges because the gain in size after fusion improves the chances of survival of the resulting chimeric individual.