Drug shortages occur frequently and are often caused by supply chain disruptions. For improvements to occur, it is necessary to be able to estimate the vulnerability of pharmaceutical supply chains. In this work, we present the first model of pharmaceutical supply chain reliability. We consider three key approaches that companies may use to improve reliability: configuration, risk of disruptions, and speed of recovery. Key metrics include expected drug shortages, average time-to-shortage, and average time-to-recovery. We parametrize the model using data from major drug shortage databases and a case example of a generic injectable oncology drug. With a lean supply chain configuration, we observe that expected shortages at status quo conditions are 10%. By either doubling the speed of recovery or halving the disruption rate, expected shortages could drop to 5%. The most influential single change would be to add a back-up supplier to a lean configuration, leading to expected shortages of 4%. We also consider profitability and present the breakeven prices for different configurations. The results from our analyses could lead to immediate policy impact, providing evidence of the benefits of redundancy and improving facility quality.