This paper presents an algorithm based on Markov chain Monte Carlo and importance sampling(MCMCIS) for
tracking multi-target in a dense environment. The joint associated events are sampled by the Markov chain Monte Carlo
and the marginal association probability of the measurement to the target is calculated. The probabilistic density is utilized
when sampling the associated events as to improve the efficiency. Although the joint probabilistic data association(JPDA) is
NP-hard, the MCMCIS provides the ability to track multi-target timely in a dense environment. The simulation experiments
are implemented to analyze the tracking precision and processing time, which shows the effectiveness of the algorithm.