In May 1982, we investigated the mortality of pre-spawning American shad (Alosa sapidissima) over a 5-hour period after passage through the 17-megawatt Kaplan turbine at Holyoke Dam, Connecticut River, MA. Radio telemetry was used to determine the survival of 36 test fish during 7 experiments by comparing their movement patterns with those of 21 sacrificed fish that were also passed through the turbine. Sixty-nine control fish fitted with dummy tags were released and held in an instream net for direct observation of mortality due to handling, tagging, and introduction procedures. The mean turbine mortality was 21.5% (95% confidence limits of 3.3-36.2%). Similar preliminary experiments with post-spawned American shad indicated that mortalities during their normal outmigration should be higher than the mortality estimate for prespawned fish.