Historical narratives typically associate financial crises with credit expansions and asset price misalignments. The question is whether some combination of measures of credit and asset prices can be used to predict these events. Borio and Lowe (2002) answer this question in the affirmative for a sample of 34 countries, but the question is surprisingly difficult to answer for individual developed countries that have faced very few, if any, financial crises in the past. To circumvent this problem, we focus on financial stress and ask whether credit and asset price movements can help predict it. To measure financial stress, we use the Financial Stress Index (FSI) developed by Illing and Liu (2006). Other innovations include the estimation and forecasting using both linear and endogenous threshold models, and a wide range of asset prices (stock and housing prices, for example). The exercise is performed for Canada, but the methodology is suitable for any country that fits the above description.