In survival regression analysis with time-dependent covariates, the covariates may be truncated/censored and may be measured with errors in some applications. In this case, a joint model for the longitudinal covariate data and the survival data is often considered to address covariate censoring and measurement errors. Typically, an empirical linear (mixed) model is assumed for the time-dependent covariates based on the observed data. However, such an empirical linear covariate model may be inappropriate for the (unobserved) truncated/censored covariate values, since the truncated covariate process may behave quite differently than the observed covariate process. In some applications such as HIV/AIDS studies, a mechanistic nonlinear model can be derived for the covariates based on a underlying data-generation mechanisms. Such a mechanistic nonlinear covariate model may provide better "predictions" for the truncated and mis-measured covariate values than empirical linear covariate model based on observed data. In this talk, we propose a joint Cox and nonlinear mixed effects model to model survival data with truncated and mis-measured time-varying covariates.