The present chapter is a systematic review describing and analyzing 85 mathematical models used for diabetes population dynamics during the last 25 years. Although diabetes is not a transmitted disease, its evolution goes through different stages, including pre-diabetes, diagnosed an undiagnosed diabetes without complications, and diabetes with treated and not treated complications. The main purpose of modeling the diabetes population dynamics is to show how the prevalence of this disease can be controlled and the complications avoided or at least limited. The models selected in this systematic review used different mathematical tools such as delay, ordinary and partial differential equations, discrete and numerical approximations, fractional-order approaches, optimal control and stochastic methods. The quasi totality of authors dealing with this theme advise health decision-makers by suggesting efficient strategies susceptible to limit the highly increasing trend of diabetes incidence and to improve the well-being of people with diabetes, essentially by avoiding the occurrence of complications.