Prescription claims have been increasingly used in drug safety and effectiveness studies to determine individual drug exposures. However, exposures based on prescriptions do not represent actual drug intake. This is known as misclassification which results in biased inference in naive analyses. We developed methods specifically tailored to correct for misclassification in the analyses of time-varying prescription-based exposures. The approach relies on a pragmatic adaptation of the misclassification simulation-extrapolation (MC-SIMEX) method. Simulation results indicate a substantial reduction bias, and this for various exposure metrics that might be of interest in different applications. In conclusion, accounting for misclassification in time-varying prescription-based exposures is challenging, but crucial to avoid biased analyses. Pragmatic methods, such as the one proposed, are a promising avenue to more appropriately account for complex types of misclassification.