Objective: To study the importance of chemical shift induced boundary artefact (CSA) at different field strengths and the implications for pulse sequence design with respect to receiver bandwidth (BW), field-of-view (FOV) and matrix size. Materials and methods: A fat-water phantom was examined in MR systems of different field strength (1.5 T, 1.0 T and 0.2 T), using pulse sequences with different receiver BW, FOV, and matrix size. The chemical shift was quantified by measuring the width of the bright and dark misregistration rims seen at the planar fat-water interface. The measured chemical shift was compared with the theoretically calculated chemical shift. Results: Excellent correlations were found between predicted chemical shift and measurement results in our experiments. The width of the CSA (in millimetres) is directly proportional to field strength, inversely proportional to receiver BW and hence to the strength of the readout gradient, directly proportional to FOV, and inversely proportional to matrix size. Conclusion: CSA occurs at all magnetic field strengths, but given a certain BW it is more pronounced at higher fields. Although the CSA in Hz is directly proportional to field strength, the visible CSA at low-field was slightly higher than theoretically expected. The relative lack of CSA in low-field strength images permits the application of narrow receiver BW sequences, resulting in increased signal to noise ratio.