The operational efficiency of transport maintenance depots is critical for infrastructure sustainability, yet systematic, data-driven evaluations in developing contexts are scarce. Existing assessments often lack longitudinal rigour and robust econometric foundations. This work aims to develop and apply a panel-data methodology for quantifying technical efficiency gains within depot systems. It seeks to identify key operational drivers and provide a replicable framework for performance benchmarking. A two-stage analytical framework is employed. First, a panel stochastic frontier model, $\ln(Output_{it}) = \beta \ln(Input_{it}) + v_{it} - u_{it}$, estimates time-varying technical efficiency for a balanced panel of depots. Second, a fixed-effects regression analyses determinants of efficiency, using robust standard errors clustered at the depot level. The mean technical efficiency score across the panel was estimated at 0.65, indicating significant potential for improvement. A one-standard-deviation increase in the spare parts inventory turnover ratio was associated with a 7.2 percentage point increase in technical efficiency (95% CI: 4.1, 10.3). The proposed panel-data methodology provides a robust tool for measuring efficiency dynamics, revealing substantial and persistent inefficiencies within the studied systems. Depot management should prioritise inventory management systems and implement continuous performance monitoring using panel-data indicators. Policymakers should adopt this framework for resource allocation and national benchmarking. Stochastic frontier analysis, panel data, technical efficiency, infrastructure maintenance, asset management, developing economies This paper provides a novel application of panel stochastic frontier analysis to transport maintenance depots, generating the first longitudinal efficiency estimates for such systems in the region and a new determinant analysis of inventory turnover impact.