The meticulous growth of both export and output of Bangladesh following the inception of trade liberalization and export promotion in the late 1970s through 2010s gravely underscore the export-led growth (ELG) hypothesis for the economy. The present study examines both the short-run causality and long-run dynamics between export and economic growth in order to confirm whether the ELG hypothesis works for Bangladesh drawing data for the period from 1974 to 2015. The long-run dynamics between export and output growth based on ARDL Bounds testing approach suggests that ELG is evident for Bangladesh. In order to examine the short-run causality, the study performs Toda-Yamamoto causality test along with the conventional Granger causality test. While the Granger causality analysis identifies a bidirectional causal relationship between exports and output growth, the Toda-Yamamoto approach confirms a unidirectional causality that runs from exports to output growth. A unidirectional causal relationship from export to output growth is necessary for the validity of the ELG hypothesis. However, bidirectional causality between export and output implies that growth in exports reinvigorates output growth which, in turn, reinforces export expansion and thereby further underlines the validity of the ELG hypothesis.