BACKGROUND:Very early mobilisation is performed in some stroke units and recommended in acute stroke clinical guidelines. It is unclear whether very early mobilisation independently improves outcome after stroke. OBJECTIVES:To determine the benefits and harms of very early mobilisation (commenced within 48 hours of stroke) compared with conventional care. SEARCH STRATEGY:We searched the Cochrane Stroke Group Trials Register (last searched April 2008). In addition, we searched 25 databases including the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 3, 2007), MEDLINE (1950 to August 2007), EMBASE (1980 to September 2007), CINAHL (1982 to December 2006), and AMED (1985 to January 2007). We also searched relevant ongoing trials and research registers (searched January 2007) and the Chinese medical database Wanfangdata (searched March 2007), handsearched journals, searched reference lists and contacted researchers in the field. SELECTION CRITERIA:Unconfounded RCTs of acute stroke patients, comparing an intervention group that started out of bed mobilisation within 48 hours of stroke and aimed to reduce time to first mobilisation and/or increase the amount or frequency (or both) of mobilisation, with conventional care. DATA COLLECTION AND ANALYSIS:One review author eliminated obviously irrelevant records; two review authors independently applied selection criteria to remaining studies. The primary outcome was death or poor outcome (dependency or institutionalisation) at the end of scheduled follow up. Secondary outcomes included mortality, dependency, institutionalisation, activities of daily living (ADLs), quality of life, time to walking, adverse events (e.g. deep vein thrombosis) and patient mood. MAIN RESULTS:One study, involving 71 participants, was included. In this study the experimental group had earlier and more frequent mobilisation than the control group (median 18.1 hours post stroke for experimental group versus 30.8 hours control; 167 minutes of mobilisation (interquartile range (IQR) 62 to 305) during admission for experimental group versus 69 (IQR 31 to 115) minutes control). Fewer patients who received early and frequent mobilisation were dead or disabled at three months, but this was not statistically significant and the confidence intervals were wide (odds ratio (OR) 0.67, 95% confidence interval (CI) 0.25 to 1.79, P = 0.42). No significant difference on any secondary outcomes of interest were found. AUTHORS' CONCLUSIONS:We found insufficient evidence to support or refute the efficacy of routine very early mobilisation after stroke, compared with conventional care. More research is required to determine the benefits and harms of very early mobilisation after stroke.