The mortality of fish captured in fishing tournaments has commonly been estimated incorrectly and, thus far, only one account has presented an estimate of the standard error or confidence interval for tournament-associated mortality. In this article we describe methods for estimating the initial, delayed, and total mortality of tournament-caught fish and provide formulae for estimating the sampling variances of these estimates. The absence of such estimates from previous studies may explain an observed lack of change in tournament-associated mortality between the 1980s and the 1990s. Our methods provide insight into the design of studies of tournament-associated mortality and suggest, for example, that many previous studies have held too few fish for observation and have greatly undersampled control fish. Improved study design and reporting should increase our understanding of the factors influencing tournament-associated mortality.