A small and light-weight wearable electrocardiograph (ECG) device with three accelerometers (x, y, and z axis) was developed for prolonged monitoring of everyday stress. It consists of an amplifier, a microcomputer with an analog/digital converter, a ti-axial accelerometer, and a memory card. Four parameters can be sampled at 1 kHz for a maximum of 27 h with the default battery and a 1 GB memory card. An algorithm for the reliable and clear detection of R wave peaks of ECG was also developed for accurate heart rate variability (HRV) analysis. The algorithm reduces ECG motion artifacts induced by body movements. Off-line data processing includes autonomic nervous system (ANS) activity bi-spectral analysis and the application of the tone-entropy method to HRV data. The availability of the system was tested through simulated office work and three-day monitoring. Both short-term and circadian rhythms of ANS activity were clearly observed. The experimental data verifies the functionality of the proposed system.