Transport maintenance depots play a critical role in ensuring efficient operation of transport systems in Tanzania. However, their effectiveness and cost-efficiency vary significantly across different depots. A randomized field trial was conducted among five randomly selected depots. Depot performance metrics were collected over a year, and statistical models were used to analyse data for cost-efficiency assessment. The analysis revealed that Depot A achieved the lowest maintenance costs per vehicle while achieving comparable service quality compared to other depots (p < 0.05). The randomized field trial provided valuable insights into the operational methodologies of transport maintenance depots in Tanzania, contributing to more cost-effective and efficient depot systems. Future research should focus on replicating this study across a wider range of depots and incorporating predictive models to forecast future costs and efficiencies. The maintenance outcome was modelled as $Y_{it}=\beta_0+\beta_1X_{it}+u_i+\varepsilon_{it}$, with robustness checked using heteroskedasticity-consistent errors.