168 Background: Oncology is a very rapidly emerging science with need for more investigational drugs. It is highly essential to accrue patients into clinical trials to develop new drugs. According to the available data, about 38 % of the CTEP/NCI supporting oncology trials close due to inadequate accrual. In this study, we propose to develop a model that will predict clinical trial accrual and retrospectively validate the effectiveness of this model. Methods: Eligible studies included treatment and supportive care intervention studies either open or closed to accrual between 09/1999 and 012/2015 at our center. Data abstracted includes coordinating group, sponsor type, presence of competing trials, trial phase, disease category, single institutional study, author, primary purpose of the study, interventional modality, targeted therapy, advanced disease, randomization, presence of placebo, rare cancer category and new investigational agents. Statistical analysis was performed in the statistic software R (R Foundation for Statistical Computing, Vienna, Austria). We used Chi-square test and Fisher’s exact test. The significance level was set to alpha = 0.05. Results: Retrospective univariate analysis at our institution showed that the variables significantly associated with clinical accrual are: the type of sponsor, author of the trial, and the type of interventional modality. Using these variables, that were statistically significant in accrual of clinical trials, we then determined the strength of our model in measuring the performance of the predictors. Our prediction model had an AUC of 64.1 %. Conclusions: Clinical trial accrual is a key factor in the development of new drugs in Oncology. Our intention was to predict factors that may affect clinical trial accrual and develop a predictive model based on it. In the future, we want to use our predictive model in a prospective validation setting in order to guide our clinical trials portfolio. Data from multiple institutions will improve the validity of this model.