This study investigates the nexus between research and development expenditure and productivity growth in Australian broadacre agriculture using country-level time-series data for the period 1953 to 2009. Using standard time-series econometrics data are analysed to examine the dynamic relationships between research and development expenditure (R&D) and total factor productivity (TFP) growth. Findings here provide econometric evidence of a co-integrating relationship between R&D and productivity growth, and a unidirectional causality emergent from R&D to TFP growth. Moreover, employing variance decomposition and impulse response function the dynamic properties of the model are explored beyond the sample periods. Findings suggest that R&D can be readily linked to the variation in productivity growth beyond the sample periods. Further, forecasting result suggests a significant out-of-sample relationship exists between the public R&D and productivity in broadacre agriculture. We used a novel method MIRR which is conceptually superior than the conventional IRR to obtain a credible estimate of returns on public research investment. We found MIRR of 10.06% per year for the reinvestment rate of 3% per year. Therefore, results establishing long run relationship between productivity and R&D in Australian agriculture shed light on the future policies in R&D investments in Australia.