In order to explore the effect of the reciprocating cutter's working parameters on cutting characteristics of cotton stalk as well as to optimize the working parameters of the reciprocating cutter, the cutting tests of cotton stalk under the double support condition were performed with a self-developed cutting test bench of cotton stalk. The samples were the ripe cotton stalks with the diameter of 11.7-13.2 mm and the moisture content of 31%-37%. The main test equipment included the self-developed cutting test bench of cotton stalk, a high-speed camera system and a digital camera. The self-developed cutting test bench of cotton stalk consisted of 3 parts: A header, a feeding device and a measurement control system. The cutting tests were grouped into the single factor tests, the response surface tests and the verifying tests. The testing factors were average cutting velocity (ranging from 0.6-1.0 m/s), cutting inclination angle (ranging from 3°-23°) and ratio of cutting velocity to feeding velocity (ranging from 1-2), and 5 levels was set for each factor. The response surface testing scheme designed with the central composite design (CCD) method was a three-factor five-level testing scheme. The response surface analysis method was used to analyze the testing dates and to optimize the working parameters of the cotton stalk reciprocating cutter. The results from the tests showed that the change trends of the objective values (the maximum cutting force per unit diameter and the cutting power per unit cutting area) with the changes of the testing factors were basically consistent with the results of the single factor tests. The effects of the average cutting velocity and the cutting inclination angle on the objective values were significantly different (P<0.01), and the effect of the ratio of cutting velocity to feeding velocity on the objective values was not significantly obvious (P>0.05), but its effect on the morphological structure of the cutting section of the stubble was significantly different. When the ratio of cutting velocity to feeding velocity ranged from 1.25 to 1.5, it was observed that the cotton stalk could be completely cut off by means of the high-speed camera system ant the cutting section was even and smooth, therefore, the quality of the cutting section was the best in the ratio range of from 1.25 to 1.5. It was concluded that the optimal combination of the working parameters of the reciprocating cutter for cutting cotton stalk was as follows: the average cutting velocity of 0.9 m/s, the cutting inclination angle of 12.7°, the ratio of cutting velocity to feeding velocity ranging from 1.25 to 1.5. According to the results from the response surface tests, a regression model was established, which described the cutting characteristics of cotton stalk under the double supports. On the basis of the results of the verifying tests, it was found that the results from the regression model were well consistent with the optimal results obtained from the response surface tests. Therefore, the regression model was fit under the experimental conditions. The results of this research can provide technical support for not only developing a cotton stalk cutter, but also designing the equipment with high efficiency and low energy consumption for harvesting hard stalk.