<p>The crucial role of green finance in environmental protection and climate change mitigation is well established in the literature. However, its effects are mostly analysed through a linear perspective, while potential heterogeneity arising from asymmetric dynamics and income thresholds is often overlooked. Symmetric ARDL models fail to capture these asymmetries and may misguide policy recommendations, particularly in contexts where long-run dynamics are crucial. By applying nonlinear ARDL (NARDL) model to a panel of 60 developing countries from 2000 to 2019, this study confirms significant nonlinear and asymmetric impacts of green finance, income and financial development on carbon emissions. Empirical results confirm that green finance exerts heterogeneous effects across different income groups, it has a statistically significant long-term mitigating effect in low- and upper-middle-income countries, but its impact is positive (i.e., emission-increasing) in lower-middle-income economies. However, the effects of green finance are significant only in the long run, underscoring the need for sustained policy commitment. Financial development and economic growth also exhibit income-dependent asymmetries in their environmental effects. These findings underscore that the effectiveness of green finance is highly contingent upon the level of economic development, requiring tailored climate finance strategies to each country’s specific developmental context. Policy implications include enhancing the targeting and design of green finance, providing subsidies for clean technologies, and implementing complementary financial and institutional reforms.</p>