There are a number of different approaches to physiotherapy treatment following stroke that, broadly speaking, are based on neurophysiological, motor learning and orthopaedic principles. Some physiotherapists base their treatment on a single approach, while others use a mixture of components from a number of different approaches. To determine if there is a difference in the recovery of postural control and lower limb function in patients with stroke if physiotherapy treatment is based on orthopaedic or neurophysiological or motor learning principles, or on a mixture of these treatment principles. We searched the Cochrane Stroke Group Trials Register (last searched May 2005), the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library Issue 2, 2005), MEDLINE (1966 to May 2005), EMBASE (1980 to May 2005) and CINAHL (1982 to May 2005). We contacted experts and researchers with an interest in stroke rehabilitation. Randomised or quasi-randomised controlled trials of physiotherapy treatment approaches aimed at promoting the recovery of postural control and lower limb function in adult participants with a clinical diagnosis of stroke. Outcomes included measures of disability, motor impairment or participation. Two review authors independently categorised the identified trials according to the inclusion and exclusion criteria, documented their methodological quality, and extracted the data. Twenty-one trials were included in the review, five of which were included in two comparisons. Eight trials compared a neurophysiological approach with another approach; eight compared a motor learning approach with another approach; and eight compared a mixed approach with another approach. A mixed approach was significantly more effective than no treatment or placebo control for improving functional independence (standardised mean difference (SMD) 0.94, 95% confidence intervals (CI) 0.08 to 1.80). There was no significant evidence that any single approach had a better outcome than any other single approach or no treatment control. There is evidence that physiotherapy intervention, using a mix of components from different approaches, is significantly more effective than no treatment or placebo control in the recovery of functional independence following stroke. There is insufficient evidence to conclude that any one physiotherapy approach is more effective in promoting recovery of lower limb function or postural control following stroke than any other approach. We recommend that future research should concentrate on investigating the effectiveness of clearly described individual techniques and task-specific treatments, regardless of their historical or philsophical origin.