Many situations require us to distribute rewards and resources between different members of a group. The members are often described with respect to their age, sex, experience, personality, and productivity. Distribution of rewards and resources requires that all these qualitatively different inputs be properly considered. How are multi-dimensional inputs handled in allocations of one-dimensional reward or resource? In Experiments 1-4, the allocation tasks had heterogeneous inputs such as effort on job and actual performance or behaviour toward administration and actual performance and allocators were college students and professional managers. All the four experiments indicated that allocators calculate separate equity ratios for each input of the allocation task and then average the two ratios in determination of reward. The commonly held belief that all the inputs are converted to a common currency of deservingness and then reward is allocated did not receive any support at all. In Experiments 5-8, the allocation task consisted of performance over two periods and the allocators were college students, professional managers, and union leaders. While college students and professional managers converted the two-year performance in a unitary measure of deservingness as is commonly believed, union people followed the model employed in Experiments 1-4. The instructions to divide fairly and to minimize conflict between the two claimants affected the perception of inputs and not the rule of reward distribution. Implications of these findings were discussed for management of reward systems in Indian organizations.