Although the conventional indirect calorimeter is a valuable tool, its size and expense prohibits its widespread use in hospitals. Furthermore, its flow-through measurement technique dilutes the respiratory variations, and hence some form of high-precision detection instrumentation is required. These limitations may be overcome by combining MEMS with CMOS circuit design technology to develop an innovative SOC biochip as the basis of a miniaturized energy consumption measurement system. In order to measure the characteristics of the oxygen sensors which form one part of this system, this study develops an automated oxygen concentration control and measurement system. This system can simulate the miniscule respiratory variations of a premature infant and can subsequently establish a suitable oxygen concentration environment to ensure the infant's well being. The proposed system is capable of establishing environments with oxygen concentrations ranging from 5% to 100%, and can control the oxygen concentration to a resolution of 0.006%. The minimum time required to increase the oxygen concentration from 21% to 100% is approximately 5.6 seconds. The proposed system can also automatically measure the properties of the oxygen sensors, including their resistance characteristics at different oxygen concentrations, the relationship between their sensitivity and the oxygen concentration, and the influence of working temperature and humidity upon their sensitivity. The necessary measurement data is acquired locally and can then be transmitted to a remote PC via the Internet.