Background: Intention-to-treat (ITT) is the standard data analysis method which includes all patients regardless of receiving treatment. Therefore, it is also a counter-intuitive type of analysis as it counts patients who did not receive treatment. Per protocol (PP) analysis includes only patients receiving treatment, but is considered to provide biased estimates. Instrumental variable (IV) analysis has been proposed as a technique to control for bias when using PP data. Objectives: To correct for bias by applying IV analysis methods using PP data from previously published individual patient data meta-analysis (IPDMA) (JCO2005;23:5074). Methods: We used center prescribing preference as an IV to assess the effects of methotrexate (MTX) in preventing chronic graft-versus-host-disease (cGVHD) in patients receiving stem cell (PBSCT) or bone marrow (BMT) transplant in 9 randomized controlled trials (1107 patients). IV methods are applied using 2-stage logistic, Probit and GMM models. Results: ITT analysis showed a statistically significant harmful effect with the use of day 11 MTX resulting in odds ratio (OR) of 1.34 (95% CI 1.02 1.76). The results using the PP data showed no difference in the rates of cGVHD with MTX use [OR 1.31 (95%CI 0.99 1.73)]. The IV estimates also showed no difference in the rates of cGVHD with use of MTX. IV IPDMA further corrected the results toward no difference in the odds of cGVHD between PBSCT vs. BMT, pointing to the possible beneficial effect of MTX in preventing cGVHD in PBSCT arm (OR 1.14; 95%CI 0.83 1.56; see table 1). Conclusion: IV estimates correct for bias and may be closer to the truth than both PP and ITT. The IV results indicate that MTX may be beneficial, contradicting the ITT analysis. To our knowledge, this is the first IV meta-analysis, which we report in the context of IPD MA.