我國經濟部自1979年積極推動科技專案，透過專家學者審查以經費補助企業從事研究發展。然而，不同產業領域的申請計畫由不同領域的專家群體審查，不同領域的專家群體在審查決策過程中可能產生決策衝突。本研究從專家群體決策角度出發，運用模糊層級分析法分析不同領域業界科技專案評審委員對於該模式的判斷，同時使用模擬來分析不同領域業界科技專案評審委員面臨不確定下對該模式的判斷變動。研究結果顯示：(1)不同領域評審委員最重視的決策準則有差異，(2)不確定性程度會影響不同領域評審委員對決策準則之重要性判斷。依據研究結果建議業界科專審查應建立決策衝突的解決機制，使得評審過程更有效率以及更透明化與公開化。 Taiwan has launched various government-sponsored technology development programs since 1979 and has provided funding for firms to do R&D. The decision of government-sponsored technology development projects is normally made by a review committee with experts from academia, industry, and the government. It is important for project selection decision to analyze different expert judgments. In this study, we applied a fuzzy analytic hierarchy process approach to comparing judgments of experts from four committees and completed a simulation to analyze changes of criteria weights under different degree of uncertainty. Our results indicated that: (1) the ranking of evaluated criteria in this study shows different among four technical committees; and (2) the relative importance of evaluated criteria is changed when experts of technical committees face different degree of uncertainty. Based on results of this study, we suggest an approach to solve conflict of expert judgments in government-sponsored R&D project selection.