Sonication is widely used for disruption of suspended soil aggregates. Calorimetric calibration allows for determining sonication power and applied energy as a measure for aggregate disrupting forces. Yet other properties of sonication devices (e.g., oscillation frequency and amplitude, sonotrode diameter) as well as procedure details (soil‐to‐water ratio, size, shape, and volume of used containers) may influence the extent of aggregate disruption in addition to the applied energy. In this study, we tested potential bias in aggregate disruption when different devices or procedures are used in laboratory routines. In nine laboratories, three reference soil samples were sonicated at 30 J mL−1 and 400 J mL−1. Aggregate disruption was estimated based on particle size distribution before and after sonication. Size distribution was obtained by standardized submerged sieving for particle size classes 2000–200 and 200–63 µm, and by dynamic imaging for particles < 63 µm. Despite differences in sonication devices and protocols used by the participants, only 16 in 216 tests of samples of the size fractions 2000–200 and 200–63 µm were identified as outliers. For the size fraction < 63 µm, fewer outliers were detected (8 in 324 tests). Four out of nine laboratories produced more than two outliers. In these laboratories, sonication devices differed from the others regarding oscillation frequencies (24 or 30 kHz compared to 20 kHz), sonotrode diameters (10 and 14 mm compared to 13 mm), and sonication power (16 W compared to > 45 W). Thus, these sonication device properties need to be listed when reporting on sonication‐based soil aggregate disruption. The overall small differences in the degree of disruption of soil aggregates between different laboratories demonstrate that sonication with the energies tested (30 and 400 J mL−1) provides replicable results despite the variations regarding procedures and equipment.