Optimal skew models explain reproductive sharing within social groups as resulting from reproductive incentives given by controlling dominants to subordinates in return for peaceful cooperation. We explore two versions of an alternative, the incomplete control model, for the evolution of reproductive sharing within groups. In this model, dominants have only limited control over the allocation of reproduction and must expend effort to increase their share of the total group output We show that, when the relatedness between dominant and subordinate is symmetrical, (1) the subordinate's fraction of reproduction either increases with, or is insensitive to, the subordinate's genetic relatedness, r, to the dominant in both versions of the incomplete control model, whereas the subordinate's fraction of reproduction decreases with increasing rin the optimal skew model, (2) the subordinate's share of reproduction in the incomplete control model must exceed that in the optimal skew model, and (3) ecological factors affecting solitary breeding success do not directly affect the subordinate's share of reproduction in incomplete control model but do in the optimal skew model. When dominant-subordinate relatedness is asymmetrical (as is often the case in parent-offspring associations), the incomplete control model predicts no reproduction by the subordinate offspring regardless of group size for groups containing any mixture of unrelated and full-sibling subordinates, whereas the optimal skew models predict that such reproduction is possible when the group size is three or more. The available evidence indicates a negative relationship between relatedness and a subordinate's reproductive share in both vertebrate and hymenopteran societies, apparently supporting the predictions of the optimal skew not incomplete control, class of models. However, such a negative relationship is not necessarily inconsistent with the incomplete control model when, as is true for some vertebrate studies, it results from a comparison of skews in genetically monogamous, nonincestuous groups of parents and their offspring (asymmetric relatednesses) with skews in groups of nonkin (symmetric relatednesses). Both models predict higher skews in parent-offspring associations. Occasional reproduction by subordinate offspring in groups of asymmetrical relatedness when such groups are larger than dyads is more consistent with the optimal skew model, however. Overall, current data on reproductive skew and its relationships to intragroup aggression and ecological constraints support the optimal skew model, but more data are needed to rule out the incomplete control model. These models are examples of two different general views of intrasocietal evolution: the tug-of-war view, in which group members engage in a struggle over resources, and the transactional view, in which group members exchange parcels of reproduction to induce beneficial behavior from each other.