The effect of the presence and growth of spoilage bacteria on the keeping quality of perishable products was modelled on theoretical premises. Growth was modelled as simply as possible, to direct attention to effects on quality and to detect how quality could be affected by the bacteria. Two approaches were investigated. Data on quality behaviour, obtained by a Monte Carlo type of simulation, based on the full model description in differential equations, were analysed with simplification of the model formulation, one mathematical and one mechanistic. Although a perfect fit with multivariate nonlinear regression was not obtained, the results of the analyses indicate practical solutions to be possible.