In order to real-time detect crop nutrition content, a smart apparatus for detecting crop nutrition was developed based on NIR spectroscopy, electronics and ZigBee wireless communication technology. It was designed to work as a wireless sensor network with one control unit and several optical sensor nodes. The incident sunlight and reflective light from crop canopy were measured at the wavebands of 550 nm, 650 nm, 766 nm and 850 nm respectively, and then the canopy reflectance and some vegetative indices were calculated. The field experiments and analyses on crop growth monitoring in wheat fields were conducted. Results showed that the correlation coefficient between NDVI and chlorophyll content was higher than RVI. The chlorophyll content and the vegetation index which was calculated with the reflectance at 550 nm and 766 nm had the highest correlation coefficients (r was 0.693 9), and the prediction model of chlorophyll content was built. The r2 of the calibration model reached to 0.494, and r2 of the predicted model reached to 0.478.