Systematic literature search for epidemiological evidence for an association of short sleep with weight gain and eventual development of obesity provided 71 original studies and seven reviews of various subsets of these studies. We have summarized the evidence for such an association with particular emphasis on prospective studies. The studies showed that short sleep duration is consistently associated with development of obesity in children and young adults, but not consistently so in older adults. We have identified critical aspects of the evidence, and assessed the possibility for interpretation of the evidence in terms of causality. We have discussed the requirement of temporal sequence between putative exposure and outcome and the implications of the time lag between them, the problems in adequate measurements of exposure and effects, the possible bidirectional causal effects, the necessary distinction between confounders and mediators, the possible confounding by weight history, and the possibility of common or upstream underlying causes. In conclusion, causal interpretation of the association is hampered by fundamental conceptual and methodological problems. Experimental studies may elucidate mechanisms, but adequate coverage of the entire pathway from sleep curtailment through obesity development is not feasible. Randomized trials are needed to assess the value of targeted interventions.