It is claimed that regular practise of transcendental meditation (TM) improves cognitive function and increases intelligence. This systematic review assesses the evidence from randomised controlled trials for cumulative effects of TM on cognitive function. Searches were made of electronic databases and the collected papers and official websites of the TM organisation. Only RCTs with objective outcome measures of the cumulative effects of TM on cognitive function were included. Trials that measured only acute effects of TM or used only neurophysiological outcome measures were excluded. 107 articles reporting the effects of TM on cognitive function were identified and 10 met the inclusion criteria. Most were excluded because they used no controls or did not randomise subjects between interventions. Of the 10 trials included, four reported large positive effects of TM on cognitive function, four were completely negative and two were largely negative in outcome. All four positive trials recruited subjects from among people favourably predisposed towards TM, and used passive control procedures. The other six trials recruited subjects with no specific interest in TM, and five of them used structured control procedures. The association observed between positive outcome, subject selection procedure and control procedure suggests that the large positive effects reported in four trials result from an expectation effect. The claim that TM has a specific and cumulative effect on cognitive function is not supported by the evidence from randomised controlled trials.