Optimal policy recommendations from natural resource management models can involve fast-changing management interventions. Actual policies often change more gradually, potentially reflecting costs associated with policy adjustment. We examine how including policy adjustment costs changes policy recommendations in models of fishery management. Specifically, we examine cases where policy adjustment costs increase at an increasing rate with the rate of change of fishing effort levels. We examine how accounting for these costs changes optimal management strategies in static and time-varying environments. Increasing policy adjustment costs results in optimal approach paths to stationary optimal solutions that are more gradual but that rely on oscillatory convergence. In time-varying environments, including policy adjustment costs smoothes optimal management recommendations, resulting in decreased variation in effort levels on the stationary part of the optimal solution, but increased variation in optimal stock sizes. We conclude that accounting for policy adjustment costs in this way leads to management recommendations that in some ways would be simpler for managers to implement (more gradual approach paths), but in other ways more challenging (oscillatory convergence) and that can affect where environmental variability is expressed in model predictions (stock sizes versus effort levels) while not suppressing this variability altogether.