PURPOSE: To present and discuss how to address bias and interpret results from survival curves developed from non-randomized designs. DESCRIPTION: Kaplan-Meier curves, the unadjusted version of survival curves, are a well-known and a relatively easy-to-understand method for displaying time-to-event data across many disease states. While intuitive, Kaplan-Meier curves do not account for group differences that can bias group comparisons using non-randomized designs.  In order to account for some of these differences, many researchers choose to “adjust” survival curves using one of several available methods. Depending on the method used to adjust survival curves, the interpretation of the adjusted survival curve can be confusing and more importantly irrelevant to the decision maker. This workshop will use a simulated dataset to present three different methods used to adjust survival curves and discuss how each method should be interpreted as well as their relevance to decision makers. The audience will be guided in a stepwise manner through the following methods of adjustment: 1) average covariate, 2) corrected group prognosis, and 3) inverse probability weighting.  The workshop will include a detailed discussion of the 3 methods with respect to their application and interpretation across clinical and costs studies. The discussion leaders will seek audience feedback on the methods including personal experience with adjusted survival curves. This workshop is appropriate for clinicians, biostatisticians, health care decision-makers, patients and others who are interested in conducting research using survival analysis.