Abstract: Abstract: The extensive use of chemical fertilizer is considered to be one of the major pollution sources to soil and water. Variable rate fertilization is an essential part of precision agriculture that can solve the ecological and environmental problems. One of the main difficulties in variable rate fertilization is the lack of the mass flow detection equipment. In this paper, an capacitance-based system for measuring fertilizer mass flow rate was designed and tested in laboratory environment. The measuring system was composed of capacitive mass flow sensor, velocity sensor and vehicle terminal. The flow sensor consisted of 3 coplanar electrodes deposited on a fertilizer pipe defining 2 capacitors: sensing and reference capacitor, and the capacitance of the sensing capacitor varied with fertilizer material. An inductive proximity switch was selected to measure the forward speed of fertilizer machine. A vehicle terminal was used to deal with the information received from the mass flow sensor and velocity sensor signals via CAN (controller area network) bus and displayed the mass flow and total mass of the fertilizer on the liquid crystal display (LCD) screen. All detected data were saved in the memory disk. In order to improve the resolution of flow sensor, the capacitive detection circuit was designed, which included a low consumption single chip microcomputer STM32F103C8T6 as the control core, a capacitive convert chip PCAP01 as the capacitance detection component, and a CAN bus transceiver TJA1050. Also, it is important to note that the sensing and reference capacitor were connected at a floating mode, ungrounded. Besides, the sensor also provided the solution for blockage detection on fertilizer machine. If the capacitance of the mass flow sensor reached the alarm threshold, an alarm was triggered and the status red LED on the sensor blinked steadily. Main interface of upper desk was developed using the Labwindows/CVI software. The advantages of this software were presented by the process of modular design, including parameters setup, mass flow and velocity information acquisition and processing, mass flow curve real-time display, data storage and replay. The influence of ambient temperature on output of the capacitance sensor was analyzed in the laboratory firstly. The data collected from the temperature experiments showed that output value of the differential capacitance sensor did not vary with temperature ranging from 15 to 55 ℃. It showed that the differential capacitance method effectively eliminated the influence of ambient temperature changes on the measurement stability. At the same time, 3 types of fertilizers (N, P, K) were used as the samples to study the relationship between fertilizer mass flow and sensor output capacitance. According to the experiment data, there was a significant linear positive correlation between the capacitance and mass flow. Then, the linear regression models were established based on the MATLAB R2008b by using the accumulated capacitance values and total mass of fertilizer. The coefficient of determination R2 was 0.988 9 for carbamideurea (N) and 0.989 8 for calcium superphosphate (P) and 0.993 5 for potassium sulfate (K). To evaluate the system's measuring accuracy, the mass flow sensor detected a series of fertilizer samples with different rotary speeds of the fertilizer bench. The results showed that the sensor had the high accuracy of fertilizer mass flow and the maximum relative error of the monitoring system was 3.75%. In addition, the experiments on pipe blockage of fertilizer machine were also carried out. The results indicated that the blockage recognition accuracy of the system was 100%, and the maximum response time was about 1.22 s. Therefore, the feasibility of detecting mass flow of fertilizer machine based on the capacitance method was verified, as well as the reliability of the linear regression model. The monitoring system designed here can help to improve performance of the variable rate fertilization control equipment.