During the past decade biocatalysis in non-conventional media has gained a lot of interest. Especially in the field of bio-organic synthesis, where poorly water-soluble substrates and products are involved, these media are very attractive.Non-conventional media generally consist of an apolar solvent phase and an aqueous phase. In this thesis, mixtures of water with water-miscible organic solvents, or water- immiscible organic solvents or (near-)supercritical solvents are described. The conventional aqueous phase contains the cellular or enzymic biocatalyst. The aqueous phase can vary from a dilute aqueous solution, with a thermodynamic water activity a w close to 1, to a dried enzyme particle with only a monolayer of adsorbed water molecules (a w < 1).In non-conventional media biocatalytic processes are governed by the presence of a phase boundary when two phases are involved. This phase boundary not only influences the rate of the bioconversion (kinetics), but also the yield of the reaction (thermodynamic equilibrium). In this thesis, several factors are described which affect the (kinetics), and thermodynamics of biocatalytic porcessen in non-conventional media.Chapter 2 gives an overview of the recent developments in the field of medium engineering for biocatalysis in non-conventional media. In this chapter a few basic design rules for the rational design are formulated. These rules may serve as useful tools for optimization of biocatalytic processes in non-conventional media.A typical example of a non-conventional reaction medium is the mixture of water and water-immiscible organic solvent. Especially for this type of reaction media the liquid-impelled loop reactor has been developed. This reactor has been used for the bioconversion of tetralin, a very toxic apolar compound. In Chapter 3 the general strategy for the selection of a suitable solvent for the bioconversion of such toxic apolar compounds in the liquid-impelled loop reactor is given, where the tetralin conversion is used as a typical example. The water-immiscible solvents should be non-toxic and nonbiodegradable. Additionally, they should reduce the toxicity of the apolar substrate and they must be practical for use in the liquid-impelled loop reactor. All the steps in the selection procedure proved to be essential. Among the 57 solvents tested, only FC-40 proofs to be suitable for bioconversion of tetralin in the liquid-impelled loop reactor. In addition, the cellular biocatalyst needs to be immobilized, to reduce emulsion formation inside the bioreactor.For the bioconversion of tetralin in the liquid-impelled loop reactor oxygen is needed. Chapter 4 describes the mass transfer of tetralin and oxygen in the liquidimpelled loop reactor from the apolar solvent phase to the aqueous phase, where the bioconversion occurs. It is found that in case of mass-transfer limitation, tetralin is the rate-limiting substrate and not oxygen.One of the selection criteria of a suitable solvent for bioconversion of apolar substates is its non-toxicity for the biocatalyst. The log Poctanol , which describes the hydrophobicity of the solvent, is a good measure for the toxicity of the solvent in a twoliquid phase system. The toxicity of a water-immiscible solvent for cellular biocatalyst is caused by two factors, i.e. the presence of a phase boundery (phase toxicity) and by the solvent molecules that are dissolved in the aqueous phase (molecular toxicity). Chapter 5 describes these effects separately. When the solvent concentration in the membrane of the cellular biocatalyst reaches a critical concentration, the solvent becomes toxic. The toxic concentration in the membrane is constant and independent of the solvent used. It is directly related via the partition coefficient over the membrane and water, to the solvent concentration in the aqueous phase. This is in turn directly related to the log Poctanol of the solvent. If the critical membrane concentration of a certain microorganism is known, the toxicity of any solvent can be predicted with thelog Poctanol .Apart from the log Poctanol , also the Hildebrand solubility parameter δcan be used as a measure of the hydrophobicity of the solvent. In Chapter 6 this parameter has been used successfully as an indicator of the solubility of apolar compounds in near-supercritical carbon dioxide (SCCO 2 ). In addition, the effect of this parameter on the transesterification rate of Lypozyme in this non-aqueous reaction medium has been studied. The change in δof near-supercritical carbon dioxide hardly influences the reaction rate. The water content of the medium influences the kinetics much more.Water not only affects the kinetics of a synthetic reaction, but it also affects the equilibrium yield of these reactions. When the thermodynamic water activity a w is decreased, water-dependent side-reactions such as in transesterification reactions are suppressed (Chapter 6). In esterification reactions, a shift in equilibrium towards synthesis is expected upon decreasing the a w .Chapter 7 describes a new method to control the a w during esterification reactions. With this a w -control method the a w can be maintained at an optimal value, at which the biocatalyst still shows sufficient activity while a high thermodynamic product yield can be obtained.This thesis actually covers two central themes in biocatalysis in non-conventional media: kinetics and thermodynamics. In Chapter 8 a general discussion highlights how thermodynamics can be used as a basic tool to reveal the processes that govern biocatalysis in non-conventional media.