Abstract: In this paper, we discuss the criteria for detecting confounders in the survival function under the assumption that the response and covariates are both continuous. By the partition-based method, we propose the necessary and suffcient conditions for a uniformly irrelevant factor when a covariate is one-dimensional. Then the suffcient conditions for a conditional uniformly irrelevant factor vector are also obtained when a covariate is multidimensional. We propose the criteria for detecting a single confounder and multiple confounders in survival function.