We conducted a retrospective analysis of data on the relationship between operatingefficiency of Kaplan turbines and direct passage survival of salmonid smolts. A review of a keyreport instrumental in establishing ±1% turbine efficiency rule for operating Snake and Columbiariver hydroelectric stations found a weak association (r 2 = 0.112) but also found misspecificationof the turbine efficiency data. At four Snake and Columbia river dams, manipulative studies wereperformed to investigate the relationship between turbine performance and smolt passagesurvival, as estimated with balloon-tag releases and recoveries. At all sites, peak passagesurvival did not coincide with the observed turbine operating efficiency peak. The differencebetween maximum survival and survival at peak turbine efficiency was as much as 3.2%.However, at three sites, maximum survival was within the ±1% peak efficiency operating rule. Ameta-analysis that used balloon-tag survival results from 11 different hydro projects also found noassociation between relative turbine efficiency at a site and smolt passage survival (r 2 = 0.0311,P = 0.2640). For the benefit of smolt survival during passage, we recommend managing turbineoperations to achieve maximum passage survival rather than focusing solely on peak operatingefficiency of Kaplan turbines.