Sleep duration has gradually diminished during the last decade while obesity and type 2 diabetes have become epidemics. Experimental sleep curtailment leads to increased appetite, hormonal disturbances and, especially, insulin resistance. Numerous epidemiological studies have therefore examined whether habitual short sleep is associated with obesity and type 2 diabetes. A large majority of cross-sectional studies have confirmed an association between short, and also long sleep duration and obesity in adults more than in the elderly. Short sleep is strongly associated to obesity in children and adolescents. Prospective studies, including studies in children, are not conclusive with regard to the effect of short sleep on the incidence of obesity. Both short and long sleep durations are associated with diabetes, but only short sleep duration seems predictive of future diabetes. Insomnia seems to be a strong contributor to short sleep duration but the association of insomnia with obesity is not clear. Insomnia is associated with type 2 diabetes and also predictive of a higher incidence. Other studies have shown that short sleep duration and insomnia are associated with, and sometime predictive of, other components of the metabolic syndrome, especially hypertension and the risk of coronary disease. The treatment of short sleep duration and insomnia with regard to their effects on the metabolic syndrome merits further study.