As the use of biotechnology products, such as genetically modified microorganisms (GMMs), in the environment might bring about undesirable ecological effects, it is important that the environmental fate of inoculant organisms, as well as any effects of their release, are assessed. Ideally, pilot studies in microcosms or small mesocosms are thus performed prior to a larger-scale (commercial) application, and the results of these studies serve to guide further environmental use of the GMM. In these pilot studies, the methods employed to assess the environmental fate and effects of the GMM will have to be fine-tuned and optimized so as to assess these phenomena in an optimal way. This review examines the methods that are currently available for the assessment of the environmental fate of genetically modified and unmodified microorganisms, as well as the impact following their release. The emphasis will be on monitoring of these phenomena in soil, as a paradigm of assessments in complex environmental matrices. Detection methods based on cultivation, which generally rely on the use of added or intrinsic markers, serve to assess the fate of the culturable fractions of released microorganisms. The specificity of this detection can be enhanced by using a combination of cultivation-based and immunology- and/or DNA-based assessments. Furthermore, specific immunofluorescence or in situ hybridization techniques are suitable to quantify populations of GMMs at the level of microscopy-detectable cells, even though fluorescent in situ hybridization is still plagued by sensitivity problems in oligotrophic environments. Detection methods based on nucleic acids (DNA or RNA) extracted from the environment offer the possibility to monitor the fate of the heterologous genes released, including that following a horizontal gene transfer. Moreover, these methods can also provide a picture of the dynamics of the total numbers of microbial cells released. For a sound assessment of the biosafety of environmental releases, a polyphasic approach to environmental monitoring is recommended, as in most analyses information will be needed on the fate of culturable and nonculturable cells, as well as on that of the specific gene sequences released. Finally, the use of impact analysis via an array of different methods is briefly reviewed, and the merits of novel approaches to assessments of community structure via molecular means (PCR/DGGE, ARDRA or T-RFLP) and metabolic profiling via Biolog are discussed as methods specifically aimed at the detection of shifts in community structure and function (metabolic complement analysis). Copyright (C) 1998 Elsevier Science B.V.