Radiation therapy has emerged as a powerful tool in the treatment of many kinds of cancers. However, the respiratory induced motion of the tumor poses significant challenges for treating tumors in the lung, breast, thorax or abdomen. This presentation addresses the issue of characterization of uncertainties associated with estimated states of the tumors motion. A classic pressure-volume dynamic model in conjunction with a pressure-displacement algebraic model is used to model the tumor dynamics. Probability density functions of the model parameters are used to express the stochastic states using a polynomial chaos series expansion. Numerical simulations are used to illustrate the accurate reproduction of the distribution of the model states as a function of time.