Over the past 20 years, poverty has conceived as a multidimensional issue, not only one-dimensional issue based on conventional indicators (i.e., income or expenditure). While a relatively huge literature has focused on the dynamic analysis of one-dimensional poverty, little attention has been given to the dynamics of multidimensional poverty. Using a panel data drawn from the Survey of Income and Living Conditions (SILC) in the years 2007-2010, this study focuses on the dynamics of multidimensional poverty in Turkey. The purposes of the study are twofold: the first is to identify “poor” in Turkey by proposing a multidimensional poverty measure that incorporates various dimensions closely related to the well-being of individuals (such as labor market, housing, health and living standards), and the second is to investigate how the new measure differs from other existing poverty measures (i.e., income poverty and EU material deprivation) by using random effect probit model. The findings show that the new measure is partially consistent with the other measures and multidimensional poverty decreased during the period under examination. Empirical work reveals that higher years of schooling, homeownership or being a rental/asset income recipient decreases the probability of being multidimensionally poor, while large household size, attachment to agricultural employment or being a social welfare income recipient increases the probability of being multidimensionally poor.