Abstract: Remote sensing driven light use efficiency models have been widely utilized to calculate the productivity of terrestrial ecosystems, and the outputs of these models are very sensitive to the values of maximum light use efficiency (εmax). In this study, city-level yield census data, MODIS reflectance data, locally observed meteorological data, and the vegetation photosynthesis model (VPM) were employed to derive annual mean city-level cropland εmax in Huang-huai-hai area (including Hebei, Shandong, and Henan) and the middle-lower Yangtze area (including Jiangsu, Anhui, Hubei, Hunan, and Jiangxi) for the period from 2001 to 2011. Then, the spatial and temporal variations of εmax and possible driving factors were analyzed. The results showed that the average annual value of the city-level means of cropland εmax in Huang-huai-hai area (including Hebei, Shandong, and Henan) and the middle-lower Yangtze area (including Jiangsu, Anhui, Hubei, Hunan, and Jiangxi) varied between 1.8-2.8 g C·MJ-1 during the period from 2001 to 2011. It was higher in the northwest and south parts and lower in the northeast and central parts. Annual mean cropland εmax increased during the period from 2001 to 2011 in most cities, but showed obvious interannual fluctuations and spatial difference in the rising amplitude. The interannual fluctuations of city-level cropland εmax were normally higher in the north than in the south and higher in the central than in the around. The annual means of cropland εmax had strong positive correlation with the amount of fertilizer used in per unit area of cultivated cropland in most cities, and it reached significant level (P<0.05), so the increase of the consumption of chemical fertilizer in these regions was one of the main causes of the increase of cropland εmax. The increase of the yield fraction of C4 crops could also induce the increase of cropland εmax. This study proves that it is of importance to develop a parameterization scheme accounting for the temporal and spatial variations of εmax for improving the calculation of productivity in croplands using light use efficiency models and remote sensing data.