Reviews of financial distress prediction models indicate that these techniques give highly reliable estimates of probabilities of default only for relatively short horizons, rarely exceeding two years. This is particularly the case when financial variables make up the sole estimate or primary estimates. So far, bank managers have focused on the one-year probability of default estimation required by Basel capital regulations. According to an emerging accounting standard (IFRS 9), banks will be obligated to estimate a probability of default lifetime in order to calculate credit allowances. Moreover, there is a need to improve communication and transparency between small and medium-sized private enterprises and suppliers of capital to overcome the problem of credit rationing, especially in Europe. Thus, it is challenging to search for new tools to extend the distress or failure prediction period. We assess the long-term (up to ten years) predictive ability of both financial and nonfinancial variables, paying special attention to the role of nonfinancial variables. Our study is based on rigorous postdevelopment distress and nondistress financial events in the Finnish environment. Our model, built with cross-sectional data from 2003, analyzes results for 2004-13 and shows that measures of solvency, turnover, environmental risk, payment behavior and board member characteristics can be significant predictors of bankruptcies for as long as ten years. Our most accurate long-range prediction results combine financial and nonfinancial variables.