The Lugeon or packer test for estimating rock permeability is not standardized and often yields anomalous results. A programme of repeated tests has been carried out in shallow drillholes in sandstones and greywackes making use of constant head tanks with continuous monitoring of flow rate and pressure measurement in the test section. These improved techniques make it possible to distinguish during tests between test system faults, such as packer leakage, and non-equilibrium effects due to the hydraulic properties of the rock mass. It is shown that the results obtained depend on the duration of the test and on the sequence in which different test pressures are applied and it is suggested that these effects are due to the presence of small but significant storage capacity in the rock mass. The relationship between equilibrium flows and pressures appears to be non-linear in both rock types.