This paper summarizes our experience with an indirect method to estimate the volume of bedload transport of coarse sediment. We report on bedload transport observations using piezoelectric bedload impact sensors (PBIS) and geophone sensors. We have performed PBIS measurements in several mountain streams. The PBIS device registers vibrations produced by bedload particles with a diameter larger than about 20 mm and records the number of impacts producing a signal above a threshold voltage during one minute intervals. We analyzed data from observations in the Erlenbach stream, where the PBIS array spans the width of an entire cross-section and is mounted flush with the surface of a large check dam immediately upstream of a sediment retention basin. To determine a calibration relationship, the PBIS data were compared with long-term sedimentation records obtained from repeated surveys of material stored in the sediment retention basin in the period from 1986 to 1999. We find that the number of impacts on the sensor per unit time is proportional to the transported sediment volume. We also assessed the relative accuracy of the method by comparing PBIS data from two rows of sensors at consecutive check dams in the Erlenbach streambed. In 1994 and 1995, we performed a calibration campaign with the PBIS system at a water intake in the Pitzbach mountain stream in Austria. There, weighing cells and a water pressure sensor in the settling basin allow sediment transport loads to be determined in 15 minute intervals. The relation between number of PBIS impulses and the sediment volumes deposited in the settling basin is approximately linear. The scatter between PBIS impulses and bedload volumes is much reduced if the measurements are averaged over one hour intervals of over daily periods. However, changing grain sizes of the transported sediment particles appear to affect the PBIS response in the Pitzbach stream. The PBIS system was tested in laboratory flume experiments using sediment particles from the Erlenbach torrent. These experiments also indicate a linear calibration relationship, and the proportionality coefficient between PBIS impulses and sediment load depends on the grain size of uniform mixtures. Overall, our investigations indicate that the relative error between predicted and measured sediment loads is within a factor of about 1.5 to 2 when using a linear calibration relation and when the sediment volume is larger than a few cubic meters for the field observations. We conclude that this non-invasive technique to continuously estimate the intensities of bedload transport in gravel bed streams can be useful for both research and practice. However, it requires calibration, which appears to depend on site-specific conditions, including factors such as flow intensity and sediment properties.