Three methods are frequently adopted to measure the permeability of granular soils: determinations made on the basis of the soil's particle size distribution; on the results of in situ rising or falling head tests; and on the results of in situ pumping tests. The paper presents the results of permeability tests made using these three methods on granular soils during investigations for the construction of the Nigg Bay fabrication yard, Cromarty Firth. Of the methods based on particle size distribution, the one due to Masch & Denny was found to give results closest to the pumping test values. Permeabilities calculated using Hazen's rule were significantly higher than the pumping test results, while those calculated from the results of in situ falling head tests were significantly lower. On the basis of the results presented, there would appear to be a case for the adoption of a method such as Masch & Denny's for permeability estimation of granular soils.