In the absence of suitable material on which to make thermal conductivity measurements, it is still possible to compute a value on the basis of an idealized model. The Maxwell model for the thermal conductivity of a two-component material, a model which has frequently been successfully used for relatively small values of porosity (<0.1) and moderate conductivity ratios (<10), may be extended to much wider ranges of these parameters (0.3 and 300, respectively) by the use of appropriate correction factors obtainable from data already in the literature.Significant errors due to overlooking a thermal contact resistance have been detected in some experimental work. The errors can be as high as 50 percent of the measured value for an air-filled sandstone of 0.2 porosity. Unless corrected, such errors can contribute a great deal to the lack of agreement between computed and experimental values. Correction for these errors also leads to much better consistency among published experimental values measured with various techniques.Conductivity values accurate to better than 15 percent may therefore be predicted for specimens saturated with low conductivity fluids such as air, steam, or oil.