A simple. rapid, reliable and highly automated analytical method has been developed to determine trimethylamine (TMA) in milk using an MS based electronic nose (SMart Nose). A set of Swedish milk samples with and without naturally occurring TMA was used to compare this new method with sensory analysis and dynamic headspace (DHS) gas chromatography with flame ionisation detection (GC-FID). The results show good agreement between these three analytical methods. A lower detection limit (0.5 mg kg(-1)) and better repeatability were achieved with this new method compared to both sensory and DHS GC-FID analyses (1-2 mg kg(-1)). The user-friendly multivariate analysis software provided with the SMart Nose system was used for TMA determination via a numerical PCA. Using Excel, peak intensity, cot-related with TMA concentration via a second order equation, gave similar results. Thus, this is an up to date method of choice for the quantitative determination of TMA in milk. It is also able to give further qualitative information on the milk analysed using the PCA analysis supplied with this equipment