ABSTRACTWe propose a new approach for calculating the dynamic range of an accelerometer based on an Allan deviation analysis of production seismic data. This test is intended as a field audit technique and does not require an unconditioned dataset from a low‐noise environment. We first show that Allan deviation can measure white noise levels using two commercial accelerometers. The analysis accurately captures the manufacturing noise density specifications and known relationships between white noise, preamplifier gain, and group forming. We then show that a production seismic dataset is suitable for an Allan deviation analysis because the results are not critically affected by a recording filter. Finally, we illustrate the proposed technique by calculating the dynamic range of an accelerometer channel in a seismic streamer using a production dataset.