The transition to a model-based forecasting environment is encountered by hurdles in a small open developing economy. An attempt is made to validate the benefits of model-based inflation forecasting for central banks in small open developing economies. Despite data limitations, two distinct VARs are designed to project near-term inflation. Batteries of tests (such as sequential forecasts, out-of-sample forecasting errors, equal-weight forecasting errors and decomposition) are performed on the two models to assess their predictive ability. The main finding is that model-based forecasts are reliable for use by central banks in small open developing economies, as substantiated by the relatively low forecasting errors.