BACKGROUND: Mobilisation of acute stroke patients--in bed and out of bed as early as possible--is currently recommended to prevent general and neurological complications. However, mobilisation protocols are poorly defined and need to be standardised in order to evaluate their clinical benefits. The purpose of this study is to highlight recent advances in recommendations for mobilisation of acute stroke and early neurological rehabilitation as found in the recent literature. METHODS: A Medline research was conducted from January 1950 to August 2005. RESULTS: Data synthesis of the observational studies and meta-analyses suggests that for most patients with acute stroke, early mobilisation is not harmful and may well be beneficial. However, no controlled study exists comparing early (e.g. 1-3 days) versus late (e.g. 1-2 weeks) mobilisation. Based on the available data and on pathophysiological considerations, we have developed a partially individualised procedure with precise entry and exit points for early mobilisation according to clinical status. CONCLUSIONS: Insufficient data are available to prove the beneficial effects of early mobilisation after stroke, although early neurological rehabilitation as part of routine stroke unit care seems to contribute to good long-term outcome. This article presents a standardised procedure for early mobilisation that may be tested in a randomised study between early mobilisation out of bed (3 days) compared to late mobilisation (6 days).