Mechanical and electronic technology seems very attractive to patients and funders as the ‘answer’ to the cost associated with disability, if robots might replace humans in giving therapy and in providing assistance, at a lower life-time cost. This hope has probably been present since Grey Walter developed an electronic robot in Bristol in 1948. However, there is little evidence to support this anticipation. Two papers in this issue report randomised controlled studies - one on the arm and one concerning walking. Large scale studies are still needed, but they do suggest that practice does indeed move someone towards perfection – by a small amount. Practice of a functional activity (task oriented therapy) does not require an accurate understanding of the specific losses causing a disability, but there are still hopes that a more focused therapy based on isolating the key loss might be more effective. The DRESS trial (unsurprisingly of dressing therapy) provides some support for a more focused therapy in people with right hemisphere cerebral damage – but a larger study is needed to confirm this. Another contrast between task practice (secondary to constraint of the good limb) and a more focused therapy (the Bobath approach) is the subject of a small trial which supports task practice as being at least equally effective. Practice is one way to learn, and learning is dependent upon feedback about performance and a study of respiratory feedback in people with chronic renal failure showed that both types of feedback studied helped when compared with no feedback. Feedback on task performance requires good measures – simple, immediate, sensitive – and two studies are published in this issue concerning mobility measures. One concerns the 200 metre fast walk test which may be more useful in cardiac rehabilitation than the six minute walk test. The other concerns a high-level mobility measure for use in neurological conditions (the HiMAT), but further comparative studies are needed to determine whether it is more useful than existing measures. Mobility limitation secondary to knee osteoarthritis is common, and arthroplasty is a common operation, and equally, ice is commonly used symptomatically after operation, but a small study fails to find any actual benefit. Rehabilitation after any loss of any skill or function ultimately aims to optimise social participation and functional adaptability, and a qualitative study here confirms that this is equally true for adults with visual loss. Finally, this issue contains a letter commenting on a published study (it is always worth reading letters to gain more insight into a study) and it also contains a series of research abstracts from a national meeting the Dutch Society for Rehabilitation which will draw your attention to other relevant research likely to be published somewhere.