Purpose This paper aims to examine whether Indonesian cross-border trade responds asymmetrically to exchange rate volatility (ERV). Design/methodology/approach An exponential generalized autorgressive conditional heteroscedasticity model is applied to estimate the ERV of Indonesia and ten main trade partners using quarterly data from 2006 to 2020. A nonlinear autoregressive distributed lag estimation is applied to estimate the impact of ERV on cross-border trade. Impacts from the global financial crisis (GFC) of 2008 and the COVID-19 pandemic are covered. Dynamic panel data is used for the robustness test. Findings In the short-run, ERV significantly affects exports to most of the top partners (positively, negatively or both). In the long run, asymmetric effects occur in Indonesia’s exports to five top destinations. The weakening of the Indonesian Rupiah mainly supports exports in the short term. Imports from top partners are also affected by ERV in both the short run and, to a lesser extent, in the long run. Both the GFC and the COVID-19 pandemic reduced trade: for most cases, in the short run. The dynamic panel model suggests that ERV has asymmetric impact on cross-border trade in the long run. Practical implications Exchange rate strategies need to avoid a single-side policy approach and, instead, account for exporter and importer differences in risk behaviour and an asymmetric response to ERV in trade. Policymakers need to consider policies that stabilise the currency. Originality/value This study provides evidence that cross-border trade can react asymmetrically to the exchange rate uncertainty and that the impacts of real ERV are asymmetric as well. The authors also apply a dynamic panel that signals that ERV matters in the long run for Indonesian trade with top partners.