During the 1960s, as more dams went into full operation on the Columbia River, it was discovered that the passage of American shad Alosa sapidissima was restricted or completely blocked through the regulating sections of some of the fish ladders. To study the problem, a full-scale laboratory model of six pools of the regulating section of the fish ladders at John Day Dam was built. American shad, chinook salmon Oncorhynchus tshawytscha, sockeye salmon O. nerka, and steelhead O. mykiss (formerly Salmo gairdneri) were counted and timed through various weir designs at different heads. American shad oriented toward surface flows, tending to reject submerged orifices as shallow as 2 m. The amount of time spent in the ladder by American shad depended on the head between pools. The species of salmonids tested were not significantly impeded by any of the weir designs or head differentials. The results of these studies were used to develop new design criteria for the regulating sections of fish ladders. These criteria have been used in modifications of existing structures and in the design of new facilities.