PM₁₀ has been identified as being a common problem in Malaysia and many other countries all over the world. A Markov chain probability model is found to fit the average daily PM₁₀ concentrations data of urban station (Shah Alam) and background area station (Jerantut) in Malaysia. This study aims to identify the occurrence of polluted and non-polluted days affected by PM₁₀ concentrations based on data for 12 years' period (2002-2013). The first order transition probability matrix of a Markov chain model and a two-state Markov chain, which are polluted days (1) and non-polluted days (0), were used for this purpose. The threshold value used in this study is referring to WHO 2006 guidelines (50µgm-3). Results of the analysis shows that there is a high probability that the next day event depends on what has happened on the previous day. The recurrence of the polluted day for Shah Alam is 4-5 days, while 2-3 days for Jerantut. By fitting the first order of Markov chain model, the results show that the higher order of Markov chain model is needed in order to get the best fitted distribution of polluted events at these two monitoring stations. Thus, the prediction of PM₁₀ concentrations event can be made by considering the conditions of the previous day event.