Objective <p>Pharmaceutical products are under constant development and improvements are continuously made. Implementing a change can have impact on the stability of the product. Comparing stability slopes of post-change batches monitored over short duration with those of pre-change batches monitored during long-term studies is not entirely straightforward as the longer the time frame the lower the uncertainty on the estimated slope, and therefore the pre-change and post-change slopes cannot be compared directly. The purpose is to develop a procedure making maximal use of all the information in the pre-change batches.</p> Methods <p>The knowledge in the complete pre-change data set in terms of slopes and variability is captured in a posterior distribution using a Bayesian analysis. Using this pre-change knowledge together with the time points used for a post-change batch, it is then studied if the observed slope of the post-change batch is in line with the predictive distribution for the post-change slope as based on the pre-change data.</p> Results <p>The proposed methodology is presented and applied to real data obtained in a stability comparability study at Merck &amp; Co., Inc., Rahway, NJ, USA.</p> Conclusion <p>The method is suitable for assessing comparability of stability slopes.</p>