ABSTRACT: Extant literature investigates on the determinants of NPL (non-performing loan) many a time, however, little is known on the relationship between economic uncertainty and NPL for any developing country. Historically, Bangladesh is suffering from high level of NPL in the banking sector. In addition to that, Bangladesh observes high level of macroeconomic uncertainty as characterized by high and volatile Gross Domestic Products and inflation, continuous exchange rate devaluations, and shallow financial markets. Motivated by this, we investigate the nexus between economic uncertainty and NPL for Bangladesh. Based on the data availability, we use the annual dataset covering 1990–2018. Furthermore, we use autoregressive distributed lag (ARDL) model considering its benefit accommodating both I(0) and I(1) variables. The empirical results show that there is positive relationship between economic uncertainty and NPL in the long run. The results are robust to the alternative measures of economic uncertainty and specifications. Consistent with the expectation, financial development carries a negative sign on the asset quality albeit the effect of economic uncertainty is actually pronounced (both the size and significance changes) in the augmented models. Key policy implication of this study is that government should formulate well anchored, rule-based policies to reduce inflation and interest related uncertainties. With a rule-based policy (for example, inflation targeting), people believe that central bank can achieve its targets. Government should adopt floating exchange rate regime that will adjust the external shocks well via exchange rates. In addition to this, government can monitor, and regulate stock market so that uncertainty in stock market is reduced. Finally, monetary and fiscal policies should be communicated well to the people. Thus, it is possible to reduce asymmetric information among the people, which may reduce economic uncertainty, and thus reduce NPL.