The emission of volatile organic compounds (VOCs) from a vinyl floor covering has been evaluated in two small climatic chambers and a microchamber using different air exchange rates and loading factors. The concentration versus time emission curves for cyclohexanone and phenol were decreasing. Evaluation of the emission data showed that a first order decay model was insufficient to describe the emission adequately. However, in spite of the different air exchange rates and chamber concentrations, the first order rate constant could be reproduced when the model was fitted to simultaneous measurements in the three chambers. This indicated that the emission was controlled by internal diffusion. A simplified model for emission controlled by internal diffusion in the source was developed, applying a diffusion coefficient which depends exponentially on the concentration in the source. This model described the emission curves satisfactorily. However, the model parameters were not reproducible, probably because the samples were inhomogeneous with respect to VOC content and effective thickness. All results indicated that the concentrations of cyclohexanone and phenol were almost homogeneous at test start (because of the wrapping of the samples during storage prior to testing) and that the emission after a short initial period became controlled by internal diffusion in the source.