Abstract: A time-discrete Markov chain procedure and a time-continuous Markov process procedure were used to study a two-stage model with time-dependent covariates for latent period of cancer. The transition probability matrix was directly parameterized by multiplication, and maximum likelihood and likelihood ratio were adopted for parameter estimation and hypothesis testing. As an example, the model was applied to analyse a set of 12 year and 6 run screening data of cervical cancer in Jingan county, Jiangxi Province, China.