Abstract: In construction scheme optimization, it is essential to comprehensively consider several random evaluation indexes, and this constitutes a problem of stochastic multiple attribute decision making. Previous studies on construction scheme optimization for high arch dams have not fully considered the random characteristics of these indexes, and therefore resulted in low accuracy in optimization. The existing methods that involve stochastic multiple attribute decision making, normally use artificial specification of the index weights and hence are not able to describe objectively the relative importance of different indexes. This paper presents a new method of construction scheme optimization for high arch dams based on stochastic dominance degrees. Through numerical simulations, we have obtained the probability distribution of construction progress indexes for all the construction schemes and calculated their stochastic dominance degrees using the stochastic dominance theory. Then, a priority order of these schemes can be achieved by calculating the weights of these indexes with CRITIC and fusing the indexes with a modified PROMETHEE II method. A case study shows that this method is effective in optimization of construction schemes for high arch dams featured with randomness in construction progress indexes.