Abstract: Effects on shock temperature measurements of the gap between the bulk metal sample and the transparent window are specially discussed in detail. A three-layer model corresponding to the bulk metal sample/gap/transparent window target-sample assembly is proposed. A universal analytical solution of the heat conduction equation of this three-layer model is obtained. The model is applied successfully to studies in radiometry shock-temperature measurements, to the analyses of the thermal resistance model and of the heat conductivity measurement of the window materials at very high pressures. Special efforts are focused on the principle and the possibility of utilizing a bulk sample rather than a film in the radiometry shock temperature measurements for metals. Based on the three-layer model, the sharp peak appearing at the beginning of the sample/window interface temperature profile will drop down at first rapidly in a few tens of nanoseconds. Then it will approach slowly to a level close to the equilibrium temperature predicted by the ideal interface model, if the width of the gap is less than 1 micron (0.5 micron less will be the better). It is therefore possible to use a lumpy sample (rather than) a film to conduct shock temperature measurement. We presented here some preliminary experimental records for the shock temperature measurement obtained by using the lumpy ironmeteorite samples, which demonstrated the feasibility of the technique. Also discussed here are the basic problems existing in the so-called thermal resistance model. Applying the analytical solution of the three-layer model according to the conditions specified by the thermal resistance model, we obtained a result contradictory both to experimental observations and to the expectation of the thermal resistance model. An insight into the thermal resistance model reveals the paradox in the basic assumptions of the model itself.