A multivariate modelling procedure is proposed in order to identify factors influencing stability, to estimate shelf-life, to select new batches for further stability testing and to evaluate changes in new batches. A model developed by the proposed procedure predicts the degradation rate constant as a function of storage temperature, pH, concentration and volume. The predicted rate constants were compared with prospectively measured rate constants, primarily from new batches stored under stress conditions, which emphasised batch differences earlier than storage under normal conditions. Strong deviations from expected rate constants led to extended testing of the batches concerned. The new data were used to upgrade the multivariate model. The procedure proposed led to the formulation of an opportunistic stability strategy (OSSY). Of the 15 batches of injectable solutions, nine batches are proposed tested by using OSSY. This led to an approximately 75% reduction in analytical measurements. Hold samples are recommended for storage under several stability conditions for back up analysis. In general, a multivariate stability model should be based on scientific data obtained from early studies, such as preformulation and formulation studies to provide both a qualitative and quantitative understanding of the mechanisms involved.