Abstract: Maintaining reviews is a core responsibility for reviewers and review groups. This paper considers the impact of updating on the results and conclusions of the Cochrane review 'Nicotine replacement therapy for smoking cessation'. This review developed from a publication in the Lancet and The Online Journal of Current Clinical Trials in 1994. It was first published in the Cochrane Library in 1996 and has had annual updates since then. The presentation describes changes made to the review; the number and type of trials included, the main estimates of effect, and the extent to which changes in the conclusions have been needed. The number of trials included in the review has increased by almost 50% since 1994, and one new type of NRT delivery system is now included. The main conclusion on the efficacy of all forms of NRT remains unchanged, the overall estimate of effect being an odds ratio of approximately 1.7. The largest change has been in the estimated effect of nicotine patch which has declined from approximately 2 in 1994 to 1.7 in 2000. We will present data exploring whether the decline in the estimate of effect for patch can be attributed to longer time to publication for trials with smaller estimated effects. Other changes to the review reflect the change in focus of research from placebo controlled efficacy trials. Many smokers now purchase NRT products over the counter without the minimal behavioural support and free drug therapy characteristic of clinical trials. Trials are now evaluating effectiveness under these conditions. Questions about the relative effect of NRT in different settings have also been addressed directly. These trials have had to be integrated with the findings from indirect comparisons using subgroup analyses. Studies making head to head comparisons with alternative pharmacotherapies have also had to be incorporated. Despite the availability of regularly up dated versions of the review, the earlier journal publication remains the most frequently cited source of evidence. However evidence for a trend toward use of the Cochrane review will be presented. Updating reviews involves more than adding new trials. The focus of interest of research may change as basic efficacy of a treatment is established. This may involve changes to the questions addressed under the objectives of a review.