A linearised version of the Cobb-Douglas stochastic frontier production function with time-varying technical efficiencies is estimated. The panel data (1990-1995) collected by the International Rice Research Institute CIRRI) from paddy farmers in the rainfed rice ecosystem of Tarlac, Philippines is used. Firm effects are assumed to be an exponential function of time. The result of the hypothesis testing indicated that significant inefficiency exists in the panel data. Empirical analysis demonstrated that paddy farms in the rainfed rice ecosystem of the Philippines have time-variant technical efficiency and this conclusion is along the lines of earlier findings in the literature. The results also depicted that technological change is not accounting for improvement in technical efficiencies which contradicts the earlier findings. Hence, empirical findings are area specific and cannot be generalised for all ecosystems and regions. On average, technical efficiency estimates improved from 80 per cent in 1990 to 91 per cent in 1995, implying that further potential that can be explored exists.