The correlation between in vitro dissolution rates and the efficiency of drug formulations establishes an opportunity for accelerated drug development. Using in silico methods to predict the dissolution rates bears the prospect of further efficiency gains by avoiding the actual synthesis of candidate formulations. Here, we present a computational protocol that achieves such prediction for molecular crystals at low undersaturation. The protocol exploits the classic spiral dissolution model to minimize the number of material parameters that require explicit molecular simulations. Comparison to available data for acetylsalicylic acid and alpha lactose monohydrate indicates a tunable accuracy within one order of magnitude. (C) 2016 Elsevier B.V. All rights reserved.