This is the first empirical evaluation of Conway’s pioneering predictions about the effects of the Green Revolution on crop yield levels, their sustainability and variability in a long-term context using a holistic approach involving economic, environmental, and ecological factors. It analyses trends in Bangladeshi rice production and identifies changing relative contributions to variations in aggregate rice output of alterations in aggregate rice yields and in the rice area cropped. Rice yields rose substantially following the Green Revolution and have been the major contributor to increasing rice output but have become almost stationary recently. This stationarity (if sustained) could result in Bangladesh finding it increasingly difficult to feed its growing population. Because of the high dependency of Bangladesh on just a few HYVs of rice (and its shrinking gene pool) the productivity of its rice crop could be vulnerable to major ecological and environmental shocks. We found that until recently, the absolute variability of rice yields was higher after the early establishment of the Green Revolution than prior to it. The relative variations in rice yields away from their trend values were smaller after the Green Revolution was well established and continued to fall with the widespread adoption of the technologies. We highlighted the trio of general factors determining rice yields. Holistic analysis requires these all to be considered. However, non-economists often overlook economic factors explored here in assessing influences on the crop yield levels while economists often do not pay adequate attention to ecological and environmental factors. Furthermore, this study contributes to the land-saving controversy involving the intensification of agriculture. The analytical framework we have employed can be adapted to other countries with similar biophysical and demographic characteristics.