Flexible parametric survival models using cubic splines become popular in survival data analysis. The property of allowing converging hazard functions leads them to be the alternatives to Cox proportional hazards model and parametric survival models. In this study, flexible parametric survival models are applied to the data set of 106 gastric cancer patients. According to this data set, metastasis and muscle contraction are found as important risk factors on survival. (C) 2017 The Authors. Published by IASE.