AbstractUsing an exploitation rate (not fishing mortality) based virtual population analysis (VPA) algorithm, which is itself a generalization of the Pope approximation to the original VPA equations, I show how to derive variance estimates for the key VPA outputs (recruitment, SSB, exploitation rates) given variance information on the key inputs (catch‐at‐age/terminal exploitation rates). Given the alternative VPA algorithm permits closed‐form solutions for the outputs in terms of the inputs, the delta method is employed to obtain the variance estimates, removing the need for complex simulation techniques. Using North Sea herring data as an example, the method’s utility is demonstrated by exploring the impact of aging error in the catch data and tuning error on the precision of estimates of SSB, recruitment and exploitation rates, and the parameters of the stock‐recruit relationship.