This article is the result of an initiative between the European Federation of Pharmaceutical Industries Associations (EFPIA) and the European Centre for the Validation of Alternative Methods (ECVAM).Its objectives are to provide the researcher in the safety evaluation laboratory with an up-to-date, easyto-use set of data sheets to aid in the study design process whilst at the same time affording maximum welfare considerations to the experimental animals.Although this article is targeted at researchers in the European Pharmaceutical Industry, it is considered that the principles underpinning the data sets and refinement proposals are equally applicable to all those who use these techniques on animals in their research, whether in research institutes, universities or other sectors of industry. The implications of this article may lead to discussion with regulators, such as those responsible for pharmacopoeial testing.There are numerous publications dealing with the administration of test substances and the removal of blood samples, and many laboratories also have their own 'in-house' guidelines that have been developed by custom and practice over many years. Within European Union Directive 86/609EEC 1 we have an obligation to refine experiments to cause the minimum amount of stress. We hope that this article will provide background data useful to those responsible for protocol design and review.This guide is based on peer-reviewed publications whenever possible, but where this is not possible we have used 'in-house' data and the experience of those on the working party (as well as helpful comments submitted by the industry) for a final opinion. The guide also addresses the continuing need to refine the techniques associated with the administration of substances and the withdrawal of blood, and suggests ways of doing so. Data-sharing between laboratories should be encouraged to avoid duplication of animal work, as well as sharing practical skills concerning animal welfare and scientific problems caused by 'overdosing' in some way or another. The recommendations in this guide refer to the 'normal' animal, and special consideration is needed, for instance, during pregnancy and lactation. Interpretation of studies may be confounded when large volumes are administered or excessive sampling employed, particularly if anaesthetics are used. Copyright The objectives of the Technical Sub group of EFPIA/ ECVAM were as follows:(i) to provide a guide on administration volumes for use in common laboratory species used in toxicity studies required by regulatory authorities; (ii) to provide consensus dosage levels for routine use that represent good practice in terms of animal welfare and practicality; (iii) to produce a guide to dosage levels representing the upper limit of common practice, which leaves scope to make the case for special investigations. Some of these suggested maximum values have been obtained from recent literature, 3,4 but appear high when compared with 'good practice' values. The need for careful attention...