Catch-curve analysis is frequently used to estimate total annual mortality (A) of exploited fishes, but the method assumes constant recruitment. Because populations of crappie Pomoxis spp. have exhibited large fluctuations in recruitment, I conducted simulations to assess the amount of variability in recruitment that precludes the use of a catch curve and compared results to recruitment dynamics in six crappie populations. Coefficients of variation (CV = 100 ?? SD/mean) in recruitment to age 1 ranged from 55% to 84% among the six crappie populations. Simulations suggested that recruitment variability in these ranges would likely allow estimation of A within ?? 10%, but the probability of obtaining estimates of A that were ??5% or more of the true A would exceed 0.15. I suggest that catch curves may be used to approximate A in crappie populations but that managers should consider the effects on management recommendations if A were ??10% of the estimated A.