Microorganisms can be used as catalysts to produce organic compounds in a highly chemo-, regio- and enantioselective manner, and whole cells do not require the costly addition of cofactors for redox reactions. However, bioconversions are slow compared to alternative chemical reactions, and the biocatalyst works at its best in an aqueous medium, while the transformations of interest frequently involve compounds with a low-aqueous solubility and that are toxic to microorganisms. This results in low-volumetric productivity in classical bioreactors. The Continuous Closed-Gas-Loop Bioreactor is described here-a reactor system with high productivity, but without the problems associated with two-phase systems, such as an emulsified product stream and phase toxicity. Its working principle is to recirculate a gas phase through a bioreaction compartment and a saturator/absorber module where the product accumulates as a clear organic solution. A wide range of bioconversions should be possible in this set-up, and proof of concept was established for the epoxidation of 1,7-octadiene to (R)-1,2-epoxyoct-7-ene by a native strain of Pseudomonas oleovorans. This reaction represents a group of terminal alkene epoxidations where the bioconversion substrate does not support growth of the microorganism. Practical results at a 51-scale are presented for this bioconversion for both batch and continuous operation with respect to the aqueous phase, showing continuous stable epoxidation at productivities >14 mu mol min(-1) L-1 (U L-1). The results confirm that the metabolism does not allow a simple optimization strategy, because growth and biotransformation substrates compete for the same enzyme sites, and conversely growth on a substrate using this very enzyme system is necessary for longterm bioconversion. Integrated removal of the CO2 formed via the liquid overflow was estimated from theory and verified in experimental work.