In 2007, 2.7 million vertebrates were used for animal experiments and other scientific purposes in Germany alone. Since 1998 there has been an increase in the number of animals used for research purposes, which is partly attributable to the growing use of transgenic animals (Fig. 1). These animals are, for instance, used as in vivo models to mimic human diseases like diabetes, cancer or Alzheimer's disease. Here, transgenic model organisms serve as valuable tools, being instrumental in facilitating the analysis of the molecular mechanisms underlying human diseases, and might contribute to the development of novel therapeutic approaches. Due to variable and, sometimes low, efficiency (depending on the species used), however, the generation of such animals often requires a large number of embryo donors and recipients. The experts evaluated methods that could possibly be utilised to reduce, refine or even replace experiments with transgenic vertebrates in the mid-term future. Among the promising alternative model organisms available at the moment are the fruit fly Drosophila melanogaster and the roundworm Caenorhabditis elegans. Specific cell culture experiments or threedimensional (3D) tissue models also offer valuable opportunities to replace experiments with transgenic animals or reduce the number of laboratory animals required by assisting in decision-making processes. Furthermore, at the workshop an in vitro technique was presented which permits the production of complete human antibodies without using genetically modified (“humanised”) animals. Up to now, genetically modified mice are widely used for this purpose. Improved breeding protocols, enhanced efficiency of mutagenesis as well as training of laboratory personnel and animal keepers can also help to reduce the numbers of laboratory animals. Well-trained staff in particular can help to minimise the pain, suffering and discomfort of animals and, at the same time, improve the quality of data obtained from animal experiments. This, in turn, can lead to a reduction in the numbers of animals needed for each experiment. The experts also came to the conclusion that the numbers of laboratory animals can be reduced by open access to a central database that provides detailed documentation of completed experiments involving transgenic animals. This documentation should not be restricted to experiments with substantial scientific results that warrant publication, but should also include those with “negative” outcome, which are usually not published. Capturing all kinds of results within such a database provides added value to the respective scientists and the scientific community as a whole; it could also help to stimulate collaborations and to ensure funding for future research. An important aspect to be considered in the generation of this kind of database is the quality and standardisation of the information provided on existing in vitro models and the respective opportunities for their use.The experts felt that the greatest potential for reducing the numbers of laboratory animals in the near future realistically might not be offered by the complete replacement of transgenic animal models but by opportunities to examine specific questions to a greater degree using in vitro models, such as cell and tissue cultures including organotypic models. The use of these models would considerably reduce the number of in vivo experiments using transgenic animals. However, the overall number of experimental animals may still be increasing or remain unaffected, e.g. when transgenic animals continue to serve as the source of primary cells and organs/tissues for in vitro experiments.