A portable coordinate measuring arm (PCMA) was a piece of portable coordinate measuring equipment that employs a series of rotating joints. In order to improve the measuring accuracy and repeatability of a PCMA, it was essential to calibrate its kinematic parameters. First, a new kinematic calibration approach for PCMAs by using a niching chaos optimization algorithm (NCOA) was proposed. A hybrid objective function for kinematic calibration was proposed that reflects the various performance tests, including the single-point articulation performance test and volumetric performance test. Then a Levenberg-Marquardt (L-M) algorithm and an NCOA are employed for calibrating the kinematic parameters. The NCOA exhibits a competitive calibration performance compared to the L-M algorithm. Experimental results show that the standard deviation of the measurement after NCOA calibration is always better than that of the L-M algorithm, and the measurement precision after calibration is improved by 40 times. An L-M algorithm and a NCOA are employed for calibrating the kinematic parameters of a PCMA. The NCOA shows better performance than the L-M algorithm.