A conventional thermal response test (TRT) provides a bulk estimate of the subsurface and borehole thermal properties over the length of the ground heat exchanger (GHE). The measurement of temperature inside the borehole during a TRT can be carried out to further determine thermal properties at different depths. The analysis of the transient temperature response is commonly performed with an analytical solution assuming a constant heat injection rate, which does not reproduce the effect of water flow along the pipe of the GHE. This heat transport mechanism can induce a temporal variation of the heat injection rate at depth although heat injection remains constant at the surface. Analysis of synthetic data generated with numerical simulations of TRTs in a layered subsurface was consequently carried out to verify this analytical approach. The program MLU was selected for analyzing the TRTs because of its capacity to take into account multiple layers. Results indicated that the analysis can be improved by accounting for variable heat injection rates determined inside the GHE. Estimation of both the subsurface thermal conductivity and the borehole thermal resistance was within 20% of the expected values, except when the thermal conductivity of the subsurface is low. For a simulation case carried out with a subsurface layer that had a thermal conductivity as low as 1 W m(-1) K-1, the borehole thermal resistance could not be determined with significant accuracy. (C) 2013 Elsevier Ltd. All rights reserved.