Sizing of ground-coupled loop heat exchangers (GLHE) depends on the ground thermal conductivity and capacity, and the borehole thermal resistance. One popular method to estimate the thermal parameters is the interpretation of in situ thermal response tests. Tile modeled response is T = (Tin + T..,)12, the average temperature of the fluid entering and leaving the ground. The Tn, response corresponds to the physically unrealistic hypothesis of constant heat flux along a borehole. Using a 3D finite element model of the borehole, we show that Tn, does not correspond to the fluid mean temperature within the borehole. Accordingly, with Tn,, an overestimation of the borehole thermal resistance results. The resistance overestimation has a noticeable economic impact. We propose instead a new estimator we name "p-linear" average of Tin and 7',,,, with parameter p --- > -1, as determined by numerical simulations. We show that the p-linear average closely fits tile average fluid temperature computed with the numerical model, hence avoiding bias in estimation of borehole thermal resistance. Finally, we discuss the problem of collinearity arising in tile estimation of thermal parameters. (c) 2008 Elsevier Ltd. All rights reserved.