The growing demand for a resource allocation model has prompted researchers in the substance use field to explore various approaches. The SDS presented by Simeone et al. (1993) was a major first step in developing a simple scale of relative needs for treatment of substance use problems that can be used in the allocation of limited resources. Its relevance for integration with ongoing treatment needs assessment surveys, as well as the emphasis placed on the need for such a scale by Federal agencies, prompted this evaluation of the SDS as a resource-allocation tool. This assessment shows that the SDS, by allocating more weight to heavily populated geographic areas regardless of their contributions to the problem, tends to exaggerate the extent of need in those areas while de-emphasizing the need in geographic areas that are sparsely populated. A scale that is less affected by population size is desirable for estimating relative treatment needs and determining resource allocation. The Relative Needs Assessment Scale proposed by the authors is more flexible than the SDS and corrects the flaws of the SDS by using weights that set a balance between the burden of the substance use problem and the size of the "population at risk" in the geographic unit. Assuming that a fairly good mix of indicators is identified for substance use problems, the alternative scale provides a better estimate of relative needs for resources. 1 table, 3 figures, and 4 references