Abstract: With a cropping area of 1.76 million hectares, the grey desert soil (hapliccalcisol) is crucial for agricultural development in Xinjiang, Northwest China. The use of chemical and manure fertilizers has been an effective way of increasing crop production. However, there has been limited information on crop yield and soil organic carbon evolution in the region under long-term fertilization using both organic and inorganic fertilizers. Thus this study was conducted to determine the relationship among carbon input, soil organic carbon and crop yield in grey desert soil in Modern Agricultural S&T Demonstration Garden of Xinjiang Academy of Agricultural Sciences. The study also analyzed the trends in variations in crop yield and soil organic carbon under different long-term fertilization practices. The results of the study will provide scientific database which can be used to improve soil fertility and promote sustainable development of crop production in the study area. A long-term experiment was conducted using various fertilizations in 1990–2013 in wheat (Triticum aestivium) and maize (Zea mays) crop rotation system in grey desert soil. The fertilization treatments included the control without fertilization (CK), chemical nitrogen plus phosphate fertilization (NP), chemical nitrogen plus phosphate and potassium fertilization (NPK), NPK plus animal manure (NPKM), two times animal manure of NPKM (hNPKM), and NPK plus straw (NPKS). The results showed that: 1) soil organic carbon contents of CK, NP and NPK decreased by 0.094 g·kg^-1·a^-1, 0.043 g·kg^-1·a^-1 and 0.053 g·kg^-1·a^-1, respectively, from 1990 to 2013. Although chemical fertilization increased crop production, but decreased soil fertility. Annual rates of soil organic carbon increase under NPKM and hNPKM treatments were 0.360 g·kg^-1 and 0.575 g·kg^-1, respectively, from 1990 to 2013. Therefore the application of manure played an important role in improvement of soil fertility for crop cultivation. 2) Compared with CK, long-term application of chemical fertilizers with manure (NPKM and hNPKM) increased crop yield. However, compared with NP and NPK treatments, chemical fertilizers with manure (NPKM and hNPKM) significantly increased wheat yield (P < 0.05). There was, however, no significant difference in maize yield between observed manure treatments and chemical fertilizers treatments (P > 0.05). The highest maize yield was 220 kg·hm^-2·a^-1 under NPK treatment. The coefficient of variation of wheat yield (29.1%–43.9%) was higher than that of maize yield (19.0%–32.7%). There were no differences between NPK with straw and NPK with manure, which suggesting that the effects of straw return on crop yield was not negligible. 3) Carbon input was significantly positively correlated with soil organic carbon and crop yield (P < 0.05). Thus increasing carbon input (manure or straw return) was a key of improving soil fertility in the grey desert soil region.