Starter cultures of lactic acid bacteria are widely used in the production of food and feed. Lactic acid bacteria are also used as probiotics in human and animal nutrition, because of a supposed beneficial influence on the intestinal flora. Preservation of large quantities of bacteria by convective drying is an economical alternative compared to freezing and freeze drying. Unfortunately, inactivation of bacteria occurs during drying. Two types of inactivation can be distinguished: thermal and dehydration inactivation. Thermal inactivation can be minimised using low temperatures or short times. Dehydration inactivation becomes the main problem in convective drying. The objective of this study was first to identify process conditions which favour the production of active dried Lactobacillus plantarum and second to determine the physical and physiological mechanisms explaining the influence of the process parameters on the dehydration inactivation.The production of dried starter cultures involves the following process steps: growth, harvest, any addition of protectant or carrier material, drying and storage. Each step involves parameters that may influence the activity of the microorganism.The drying parameters studied were drying temperature and drying time. A number of drying methods with varying drying rates were used. The drying procedure proved to have an important influence on the residual activity of the dried L. plantarum. The highest residual activity was achieved after convective drying of a layer at 30°C. The residual activities were depressed using lower drying temperatures or using drying methods with a very short or a very long drying time. These experiments confirmed the following hypothesis: during drying at 30°C a physiological adaptation of the microorganism occurs which is time dependent.The influence of cell concentration before drying was studied for cells grown under standard growth conditions as well as for cells grown with osmotic stress. The residual activity of cells grown under standard conditions was higher when high initial cell concentrations were used. Variation in initial cell concentration did not influence the residual activity of cells grown with osmotic stress.Growth of micro-organisms under stress situations can result in adaptations, which might improve the residual activity after drying. Therefore, the influence of the following growth conditions was studied on the dehydration inactivation of L. plantarum : composition of the growth medium, osmotic stress, pH-control during growth, growth phase and reactor concept. The highest activities were achieved applying optimal growth conditions. Stress during growth did not result in an improvement of the residual activity after drying.Carbohydrate addition could improve the residual activity after fluidized bed drying. There was a difference in effectivity of the carbohydrates. Sorbitol, maltose and sucrose addition resulted in increase in residual activity, whereas lactose, trehalose, glucose and fructose were not protective. In order to get a better understanding of the mechanism of protection, a physical and a physiological approach were chosen.Water activity and moisture distribution at a given overall moisture content are influenced by the addition of carbohydrates. Desorption isotherms have been measured in order to determine the moisture distribution in a mixture. Calculations showed that significant shifts may occur in water activity and moisture distribution as a result of the presence of carbohydrates. These shifts could not explain the protection. Former research showed that dehydration inactivation is associated with cell membrane damage. This damage may occur when the cell membrane undergoes a phase transition from the liquid crystalline phase to the gel phase and vice versa during drying and rehydration. Membrane phase transition temperatures were determined using Fourier transform infrared spectroscopy in order to study whether carbohydrates influence the phase transition behaviour of the membrane by interaction with the phospholipid head groups of the membrane. Surprisingly, there was only a small increase in phase transition temperature during drying, which was independent of the presence of carbohydrates. This phase behaviour could be attributed tot the molecular structure of phosphatidylglycerol and lysyl- phosphatidylglycerol, the predominant phospholipids in L. plantarum. The relatively high activity of dried L. plantarum can be attributed to its low membrane phase transition temperature.Finally the complete process was considered. The drying procedure and the use of additives were the most effective process steps in improving the activity of dried L. plantarum. The applied process oriented approach has led to a better understanding of the importance of several parameters.