Compact urban development is preferred as a central strategy for promoting sustainable urbanization worldwide. By facilitating shorter trip lengths, reduced travel demand, and promoting public transit facilities, compact urban development is widely associated with lower CO₂ emissions from transport. However, urban compactness has been mostly equated with high density, overlooking other spatial characteristics of compactness that shape travel pattern and emission outcomes. This study revisits the compact city paradigm by defining compactness through multiple spatial dimensions and examining their association with CO₂ emissions from transport in 28 Indian cities. Five dimensions of compactness—density, shape, contiguity, spatial concentration and fragmentation—are quantified using Landsat data. Spatial metrics are calculated using ArcGIS Pro and Fragstats. A two-step analytical framework is adopted to assess the relationship between dimensions of compactness and related emissions. An ordinary least squares (OLS) regression model with heteroskedasticity-consistent (HC3) standard errors examined linear effects, and a Random Forest model with leave-one-out cross-validation evaluated nonlinear patterns and verified the relative importance of each dimension. The results indicate that fragmentation is the strongest positive influencer of per-capita emissions, with spatially discontinuous urban growth considerably increasing CO₂ emissions from transport. In contrast, contiguity and spatial concentration exhibit strong negative correlations with emissions, showing that spatially aggregated development patterns reduce emissions. Density and shape compactness have negative but moderate impacts. The Random Forest analysis supports the relative importance ranking observed in the regression results. These findings imply that reducing fragmentation and strengthening contiguity and concentration could achieve greater emission reductions than increasing density alone. For rapidly urbanizing Indian cities, policies that encourage connected and concentrated compact urban developments are crucial for low-carbon mobility transitions. Future research could integrate inter-city connectivity and network centrality perspectives to better capture multi-scalar influences on transport CO₂ emissions.