Markov chain is a stochastic process to describe a phenomenon in the future based on a previous state. In practice, Markov chains are distinguished by time into two, namely discrete-time Markov chain and continuous-time Markov Chain. This research will discuss the continuous-time Markov chain with finite-state. COVID-19 phenomena can describe and predict using the continuous-time Markov chain. Authors use the data daily cases of COVID-19 in Greater Bandung including Bandung City, Bandung District, West Bandung District, Cimahi City and Sumedang District. Used data came from simulated data of daily cases of COVID-19 in Greater Bandung from August, 2020 until November 14, 2021 that recorded through the website COVID-19 of West Java. In terms of described and predicted the COVID-19 phenomenon in Greater Bandung for long-term probability, authors use stationary distribution and limit distribution. COVID-19 phenomenon is described into two states: state 0 (lower than average of data) and state 1 (higher than average of data). The result of continuous-time Markov chain with finite-state shows that the probability of the daily cases of COVID-19 for five locations in Greater Bandung is state 0 have a larger probability than state 1. It means that COVID-19 in Greater Bandung over the long-term will decrease.