Snacking is often presumed to contribute to obesity, but to date, studies have not demonstrated such a causal relationship, probably because a clear definition of snacking is still elusive. The usual one, i.e. any intake between traditional meals, has no physiological basis. Moreover, because some evidence suggests that frequent meals may prevent overweight, any confusion between snacks and meals may mask the deleterious effect of snacks on energy balance. Therefore, we developed a biobehavioral approach to assess whether objective criteria for eating a meal and snacking could be determined. Our main findings were that regardless of the time of consumption or macronutrient composition, snacks exerted a weak satiety effect, with those higher in protein having the strongest. The energy content of snacks was never compensated for at the next meal and led consistently to a positive energy balance compared with no-snack conditions. Biologically, the snack-induced insulin secretion suppressed the late increase in plasma FFA, which may have contributed to the inhibition of satiety. Lastly, snacking was not preceded by the glucose and insulin profile observed prior to a spontaneously requested meal. In conclusion, further studies on the role of snacking in energy balance should include criteria other than nutrient composition or consumption between meals for defining these eating occasions as snacks.