The electron-capture and beta(-) -emitting radionuclide 204Tl has been known from previous experience to show discrepancies between different methods of standardisation. Source preparation is also difficult due to the complex chemistry of thallium. The Bureau International des Poids et Mesures (BIPM) intercomparison held in 1997, showed discrepancies of up to 10% between measurements of solid sources and liquid scintillation methods. These problems have been the subject of a BIPM CCRI(II) working group since 1999. This paper presents the results from a primary standardisation of a 204Tl solution using 4pibeta-gamma coincidence counting and liquid scintillation counting. The tracer technique was used for the 4pibeta-gamma coincidence counting, where 204Tl was traced with 134Cs and 60Co. The extrapolation to 100% beta-efficiency was performed by three different approaches: foiling, adding carrier and measuring sources of different initial masses. The results showed that tracing with 60Co and using external foiling gave the same result as obtained by liquid scintillation counting. A comparison of the results achieved by the different methods of measurement, tracers and methods for extrapolation is presented and discussed in this paper.