The performances of Modified Atmosphere Packaging (MAP) for perishable foodstuffs depend mainly on food composition, on the initial microbiological qualities, on the nature of the packaging and on gaz contents. The expected shelf life will be limited by the bacterial development and the evolution of organoleptic properties. The solubility of the gas and diffusion according to the food are going to determine its efficiency. The permeability of the packaging will act on the partial pressures in gas. Methods were developed, data and models were acquired allowing the determination and the optimization of the efficiency of modified atmospheres packaging. Computing and analytical methods were used for the collection of data required for the establishment of mathematical models. Data on permeability’s properties of packaging were acquired from published data by methods used in engineering of the knowledge in order to integrate in a semi automatic way data from heterogeneous and multiple sources. A database on the packaging permeability data and on gaz solubility and of diffusivity data was then established. A synthesis of the available methods for the measures of the solubility and the diffusivity of O2 and CO2 was realized. The solubility in the CO2 was acquired by chemical titration and reaction with a scavenger based on barium, and that of O2 by luminescence. The diffusivity of gases is acquired by local quantification, determination of the gradients of gas and modelling. Methods of quantification of the effect of the CO2 on bacteria independently of the acidification were developed and thus the active agent was identified as the ion H2CO3. The largest concentration of CO2 that prevent the growth (CMI) were determined. The bacterial effect of CO2 was quantified by determining the minimal inhibitory concentration. For the aerobic germs, the kinetic of growth according to the content in O2 was modelled and for the anaerobic the maximal quantities allowing the growth were acquired. The laws of gases transfers were combined with bacterial growth models. The models for gases transfers were validated in the food. The effects of gases on the bacterial development on food were combined with already existing models on the effects of temperature and pH. A modelling tool was developed. The results of the project will allow the companies to optimize packaging size and composition, to reduce their volume, to guarantee shelf life without preservative and also to avoid the wasting.