Sample size re-estimation (SSR) allows a trial to change sample size based on interim results using a pre-specified criteria. It provides an opportunity to increase probability of successful trial in case that treatment shows trend of efficacy at interims, but underpowered with original sample size. In this presentation, we will review statistical methods to implement SSR, use simulations to show operating characteristics, and compare SSR with classical group sequential design. We will also present simulation results in a more complicated design where there are two endpoints and both allow SSR at interim analysis. Operational challenges of applying SSR will also be discussed.