Background: The NNT is a well-known effect measure for reporting the results of randomized controlled trials (RCTs). It represents the average number of patients who must be treated to prevent one adverse outcome within a certain duration of follow-up time. It was shown that NNTs are used in 62 of 734 articles (8%) presenting results from RCTs with binary and time-to-event outcomes in the medical literature (Hildebrandt et al. BMC Med Res Methodol 2009). However, in the case of time-to-event data incorrect methods were frequently applied. Confidence intervals for NNTs were given in 21 of 62 NNT reporting articles only. The Cochrane Handbook gives detailed information about how to use NNTs in presenting the results of reviews. However, the frequency and adequacy of the usage of NNTs in Cochrane reviews is unknown. Objectives: To investigate whether and how review authors used NNTs to present the results of Cochrane reviews. Methods: We randomly selected 100 reviews each from the years 2003, 2005 and 2007 of The Cochrane Library. If NNTs are presented we looked for the outcome for which it is calculated, whether the calculation is appropriate, and whether a confidence interval is givenwiththe estimate. Results: We found that 63 of 219 reviews (29%) and 18 of 81 protocols (22%) presented or planned to use NNTs. The percentage of Cochrane publications using NNTs is equally distributed across the years 2003 (27%), 2005 (26%) and 2007 (28%). We will present our final results at the Colloquium. Conclusions: NNTs are more frequently used in Cochrane reviews than in articles presenting results of RCTs in medical journals.