Following the accident at the nuclear power station in Chernobyl, KNMI and RIVM developed a model for the long-range atmospheric transport of radionuclides. This model, the RIVM/KNMI PUFF model, is in use for application in the National Plan for Nuclear Emergency Planning and Response. The PUFF-model is used at KNMI and RIVM. However, differences exist between the two versions of the model due to new model developments. It is therefore decided to create a new version of the PUFF model, the NPK-PUFF model. This model combines the best elements of the two versions , especially in the use of the more detailed HIRLAM- and ECMWF- meteorological data. The NPK-PUFF model is now operational at KNMI and RIVM. However, before the model can be used in the National Plan for Nuclear Emergency Planning and Response, it is necessary to check whether the two operational versions of the NPK-PUFF model give similar results. This report describes the tests performed and the results. The tests consist of a calculation for a hypothetical release on 17 November 1994, using different combinations of meteorological data, and a calculation with actual data in June 1997. Various criteria are used to compare the results and to accept or reject the results. The results of the two operational versions of the NPK-PUFF model appeared to be similar. The differences in model results were small and the deviations were not systematically. For some calculations, at a few moments the criteria were not fulfilled completely. However, these deviations are considered not to be crucial. With these results, it is proposed to accept the NPK-PUFF model, version 1.0, for application in the National Plan for Nuclear Emergency Planning and Response.