The contractile activity of shivering muscles is usually measured by electromyography using intramuscular electrodes. This methods shows good correlation to overall heat production, but is an invasive method. In addition to the electrical potentials produced by active muscle cells, the microvibrations ("myoacoustic" signals) that emanate from each contacting motor unit could, in principle, be used for measuring shivering quantitatively in a non-invasive fashion. To test this, we measured shivering activity in the pectoral muscles of Japanese quails (N=8) and pigeons (N=8) by simultaneous electromyography and piezoelectric acoustic myography. The measurements were done at three ambient temperatures (10, -5, and -20 °C) both in control and cold-acclimated birds. For electromyograms (EMG), we used three intramuscular 3-mm needle electrodes fixed in an epoxy plate in a triangular configuration. To measure acoustic myograms (AMG), we placed a piezoelectric sensor (R99–900, CliniLab, Finland, diameter 0.7 mm, length 7 mm). The signals were amplified and filtered appropriately, and fed to an 16-bit A/D converter and sampled at 1000 Hz. Root mean square (r.m.s.) values of ca. 10-s data segments were used for estimating EMG- and AMG- amplitudes. AMG and EMG showed significant correlation in all individuals and at all tested ambient temperatures: 1) The correlations between the r.m.s. values of successive EMG- and AMG-segments were significant in each individual at each temperature. 2) The overall r.m.s. values between temperatures also showed significant correlations. Furthermore, the acclimation state of the bird had no effect on the relation between EMGs and AMGs. We conclude that acoustic myography is a convenient and inexpensive method to measure shivering.