Gas and oil companies frequently encounter build up of Naturally Occurring Radioactive Material (NORM) in their production and processing facilities. In the Netherlands NORM is subject to strict national regulations and, consequently, installations have to be screened on a regular basis. The availability of accurate and reliable NORM sampling and analysis techniques is therefore essential. A number of years ago, the Nederlandse Aardolie Maatschappij B.V. (NAM) actively initiated an investigation on analysis techniques for NORM samples from gas and oil companies. Within this framework, Shell Research Amsterdam organized a four-stage interlaboratory test programme in which representative samples of increasing complexity were analyzed by a number of Dutch institutes. Whereas a large spread in results was observed in the first stage, results in the last stage deviated less than ±10% from the values certified by an independent referee institute, even for comple, sludge samples. It was found that in particular the use of different values for the -yields and branching ratios amongst the institutes was responsible for the initial spread.