Estimating the size of illicit drug users is pivotal for government to deliver public health services. However, it is often a difficult task since illicit drug users are not able to be reached by regular approaches and are often considered as hidden population. Using the dataset of “Central Registration System for Drug Abuser of Macao” collected by Social Welfare Bureau (SWB) Macau, current research aims to apply capture recapture method to provide a relatively accurate estimate of the illicit drug users in Macau. Capture-recapture method has three critical assumptions: closed population, population homogeneity and stable marking. Nevertheless, since illicit drug users in Macau merely satisfy assumptions of closed population and stable marking, the current study applied Mthb model to address population heterogeneity. Moreover, simulation study indicates that Mthb models generally produce reliable estimates if the true model indeed is heterogeneous. The final model yields an estimate of 2929 illicit drug users with a confidence interval of [2543.60, 3455.52], indicating a possible underestimate of the size of drug users from the CRSDAM annual. A potential limitation of this study is whether the estimated individual probability of being captured is accurate Future work is needed to improve the current approach, for example, besides demographics, the CRSDAM also provides information about III behavioral characteristics of drug using, which could be useful in improving the estimation of capture probability