With the cultivatable lands dwindling and concern over food security increasing in China, understanding the productivity of its croplands has become increasingly significant. Taking the drylands in the Northeast and North China Plains as examples, we estimated the production potential of maize based on spatial variation of light and temperature in these regions. Using data obtained from the national statistics bureau in a combination with land, precipitation, soil, social and economic data from 1995 to 2015, we calculated spatiotemporal variation of productivity of the cultivable lands and its determinants using the potential attenuation model. The results showed that: 1) The light-temperature-water production, light-temperature-water-soil production, light-temperature-water-soil-socio-economic production of the drylands in the Northeast China Plain were all higher than those in the North China Plain. Spatially, the productivity of the drylands in both plains increased from the north to the south. 2) Correction analysis between the determinants revealed that investments in social and economic development was extremely important for protecting production of the available cultivable lands, especially the drylands in the North China Plain that has been in degradation. 3) In both plains, the difference in water yield gap was lower in the south than in the north, and lower in the east than in the west, while the difference in soil yield gap was higher in the south than in the north; the difference in socio-economic yield gap was higher in the north than in the south; these spatial variations were closely related to the zonal geography in China, including soil quality, natural environment and other social and economic factors in the Northeast Plain. 4) The yield gaps caused by different factors in both plains were ranked in yield gap affected by socio-economic factor > yield gap affected by soil factor > yield gap affected by water factor. The difference in social and economic output was small in the drylands in the north China Plain, and the difference in soil yield was small in the drylands in the Northeast China Plain. These further proved that investment in social and economic development was most important for maintaining productivity of the cultivatable dryland in the North China Plain while the inherent natural environment was most important for ensuring productivity of the drylands in the Northeast China Plain. The potential attenuation model is an effective to estimate productivity of large-scale farmland, and its results provide guidance for sustaining high and stable crop yield while in the meanwhile reducing land degradation in different agricultural regions. [ABSTRACT FROM AUTHOR]