Subjects who scored low and high on Fiedler's least preferred coworker scale were provided with information about behavior and performance of two workers of several groups, and were asked to distribute a fixed sum of money between the two workers of each group. The principal point of interest centered around the performance of the two groups of allocators on the equity integration model. Results from four experiments showed that high scorers (relation-oriented) allocate reward according to the precise prescriptions of the equity integration model, whereas low scorers (task-oriented) vary in their performance on the task. This difference between the task- and relation-oriented persons disappeared when the allocation task consisted of task-relevant inputs such as effort and performance. This result shows that performance of the two groups of subjects is contingent upon the nature of inputs entering into the allocation task. Task-oriented persons also had a tendency to reject claim of workers with both negative inputs. Implications of these results were discussed for interpretation of Fiedler's measure and contingency theory as well as for information integration theory.