This paper uses micro-level historical data to examine the causes of bank failure. For state charactered Kansas banks during 19 10-28, time-to-failure is explicitly modeled using a proportional hazards framework. In addition to standard financial ratios, this study includes membership in the voluntary state deposit insurance system and measures of technical efficiency to explain bank failure. The results indicate that deposit insurance system membership increased theprobability of failure and banks which were technically inefficient were more likely to fail than technically efficient banks.