The pair of papers of which this is the second presents some of the commonest methods for calculating long-wave radiation heat exchange in a room. Paper I covered the theory behind the various methods. These vary in complexity from accurate calculation of view factors with multiple inter-reflections between surfaces to simple, non-geometric methods using fixed heat transfer coefficients. This paper (Paper II) presents the results of a set of tests which were developed to compare the performance of these algorithms with an analytically exact method. The main conclusion is that in many cases it is possible to use a non-geometric model without significant loss of accuracy. A more accurate method is required in cases of low surface emissivity or large temperature differences between surfaces. The importance of internal long-wave radiation for the overall energy balance of typical houses was tested, and the results are presented here. If no account is taken of long-wave exchange (disabling exchanges by setting all internal emissivities to zero), predicted energy requirements change by up to -11% in the type of house and heating system tested. Greater changes could arise in different house types or those with a larger radiant output from the heating system. The results also showed the importance of modelling long-wave exchange to and from windows explicitly.