In Chapter 4, the modelling of the time variable in longitudinal data analysis is discussed. In general, a distinction can be made between time as a continuous variable and time as a categorical variable represented by dummy variables. The latter is only possible in a longitudinal study with fixed time-points. Basically, three different situations are explained. (1) Growth curve analysis; i.e. when the time variable is the covariate of interest. Regarding growth curve analysis, the chapter also includes a discussion about latent class growth curve modelling. (2) The time variable as a potential confounder. When the time variable is a potential confounder it is argued that it is important to distinguish the time variable from age and (3) the time variable as effect modifier. When the time variable is an effect modifier, the interest is either in a different development over time for different groups or in a different relationship between an outcome variable and a covariate in different parts of the longitudinal study. Both situations are discussed in detail. All methods are accompanied by extensive real-life data examples.