Most of the large rivers of South America are impounded mainly for hydropower production. The construction of fish passes has been one of the strategies adopted by Brazilian authorities and the energy sector to diminish the effects of these barriers on migratory fish communities. Despite the high investments and efforts involved, most facilities have been considered ineffective for conservation purposes. Decades of poor monitoring and the lack of specific studies have limited our knowledge on the real role of fish passes. Efficiency has been frequently defined as the proportion of fish that ascend a fish passage facility, compared to the shoal size that reaches the dam. Inspired by the notion that fishes accumulated below the dam need to migrate upstream, the quantity of fish passed upstream historically indicated successful management, as generally inferred by the fish abundance inside the fish pass. We propose a new concept for estimating fish pass efficiency for South American rivers, based on the capability of the fish pass to maintain viable populations. This broader approach is necessary because knowledge of fish habitats below and above the pass, plus the feasibility of downstream movements of eggs, larvae and adults through the reservoir and past the dam, is needed for assessing whether a fish pass is working as a conservation tool. Copyright © 2011 John Wiley & Sons, Ltd.