For the determination of trace amounts of iridium and rhodium in photographic emulsions different sample treatment procedures were coupled with inductively coupled plasma mass spectrometry (ICP-MS) and, for iridium, also with negative thermal ionisation isotope dilution mass spectrometry (NTI-IDMS) and ICP-IDMS. IDMS determinations of iridium were carried out using an enriched 191Ir spike solution. Elimination of the silver matrix was established for both mass spectrometric methods, which prevents corresponding memory effects in the ICP-MS system and is essential for the formation of IrO2 - thermal ions, respectively. For NTI-MS measurements, the gelatine matrix of the emulsion must also be eliminated by decomposition with HNO3-H2O2. The detection limits for ICP-IDMS and NTI-IDMS are 0.1 and 10 ng iridium per gram of emulsion, respectively, so that all relevant photographic emulsions, doped with iridium, could be analysed. ICP-IDMS with silver matrix elimination is recommended as a routine method, while NTI-IDMS can be applied as an accurate reference method. Determinations of the mono-isotopic element rhodium cannot be carried out by IDMS. Photographic emulsions were therefore first dissolved in concentrated ammonia solution and rhodium was then directly measured by ICP-MS after dilution of the solution using an internal indium standard for calibration. This procedure has the advantage of a simple sample preparation technique but introduces high amounts of silver into the mass spectrometer and dilution of the sample restricts the detection limit. Because of the similar chemical and physical behaviour of Ru and Rh, ruthenium was used as an internal standard during matrix separation and subsequent ICP-MS measurements of trace amounts of rhodium. By determining the 103Rh/99Ru ratio of the separated sample, reliable results for the determination of trace amounts of rhodium were obtained, which was confirmed by the good agreement between the data obtained by the procedure with internal Ru calibration and those obtained without matrix separation and calibration by indium. The determination of trace amounts of iridium and rhodium in the range 0.5-650 ng g-1 was carried out in photographic emulsions with the different methods, which agreed well in their results. The relative standard deviations were usually in the range 3-10% and were substantially influenced by sample inhomogeneities.