To monitor the low back risk imposed by asymmetric postures at workplaces, a method using angular velocity sensors was studied. According to a simple model analysis, trunk rotation could be calculated from the angular velocities measured at both the waist and shoulder and from the inclination of each angular velocity sensor. We thus developed a new detector consisting of an angular velocity sensor (ENC-05D, Murata, Japan) for detecting angular velocity and an acceleration sensor (ADXL05, Analog Devices, USA) for measuring inclination. The precision of the angular velocity sensor was high as the correlation coefficient between the output of the sensor and the true value was 0.9996. When the detectors were affixed to a subject and compared with data measured by a Vicon System 370 (Oxford Metrics, UK), the correlation coefficients between the two methods were 0.949 and 0.815 during model tasks of box transfer and box lifting, respectively. In a model of lifting boxes at different rates, the mean and standard deviation increased according to the task speed. This method was shown to be of practical use for monitoring trunk rotation.