SummaryWe included 34 trials with 3742 participants, identified through 6 database and supplementary searches (to May 2017): 29 were randomised; 4 were quasi-randomised and 1 was cluster-randomised. Disparate measurements and outcomes precluded meta-analyses. Blinding was attempted in only 6 out of 34 (18%) trials. A multimedia format, alone or in combination with text or verbal formats, was studied in 20/34 (59%) trials: pre-operative anxiety was unaffected in 10 out of 14 trials and reduced by the multimedia format in three; postoperative anxiety was unaffected in four out of five trials in which formats were compared. Multimedia formats increased knowledge more than text, which in turn increased knowledge more than verbal formats. Other outcomes were unaffected by information format. The timing of information did not affect pre-operative anxiety, postoperative pain or length of stay. In conclusion, the effects of pre-operative information on peri-operative anxiety and other outcomes were affected little by format or timing.