The generalized linear mixed model has emerged as a routinely employed class of linear models where both fixed and random componentsRandom component are considered for analyzing follow-up data. In a mixed model, the underlying conditional distributions for given random effects need not be Gaussian. The quasi-likelihood-based linearization, penalized quasi-likelihoodQuasi likelihood, and pseudo-likelihood-based approach are included in this chapter. This chapter provides generalized linear mixed models in a coherent manner with theoretical perspectives addressed with limitations and advantages for modeling binary, count and time-to-eventTime to event data.