AIDS epidemic and HIV dynamics modeling can be traced back to 1980’s soon after the AIDS epidemic was recognized. In AIDS epidemic modeling, both stochastic and deterministic differential or difference equations have been used to describe the person-to-person transmission and the epidemic in high risk populations as well as in the general population. Some statistical and epidemiological approaches such as back-calculation have been proposed to estimate the AIDS epidemic. In modeling HIV dynamics within a host, biomathematicians and theoretical biologists have made great advances in the development of mathematical models to study the characteristics of HIV replication and HIV evolution. Most of these models are differential equations or compartmental models. A major gap in HIV dynamic modeling research is the lack of statistical methods to conduct formal and rigorous inference and analysis based on clinical or experimental data for the complicated HIV dynamic models. Our group has made and is continuing to make our best effort to develop relevant statistical inference methods for HIV dynamic models and data analysis. We are currently working on parameter estimation, identifiability problems, inference methods, model evaluation and comparison, study design, and model-based predictions of AIDS clinical outcomes with HIV dynamic models. We are also developing user-friendly software packages to perform the above tasks.