本文基於資料包絡法(dataenvelopment analysis)之架構，針對多決策單位之組織，除評估其效率外，並冀望在總資源不變之條件下，將資源重新分配與各決策單位，以期能獲得整體更高的效率。在模式構建部份，稍不同於一般之資料包絡法，將接受重新分自己之投入因子視為變數，而列出非線性規畫模式。按著引用一般化線性規畫(generalizedlinear programming)問題解法，求解出各決策單位所認為最理想之資源分配方式。由於各決策單位所認定之最佳分配方式不盡相同，因此必須以組織整體效率為考量，由各決策單位所認定之分自己方式，挑選出最理想者。文中並舉一簡例，說明此理念確實可提升組織整體之效率。

　　Based on the structure of the data envelopment analysis(DEA) approach, this paper measures the relative efficiency of an organization via resource allocation. When the total amount of a resource is fixed, it can be distributed to each unìt wìthin an organization in a better way. This problem is formulated as a nonlinear program similar to the generalized linear program(GLP). Consequently, the solution method of the GLP can be applied. Since different units may have different ways of allocating the resource, the resource, the best one in terms of aggregate efficiency is selected. A simple example is used to illustrate this idea of resource allocation. The results show that the aggregate efficiency of an organization can really be improved.