Disturbed sleep may negatively influence physical health, cognitive performance, metabolism, and general wellbeing. Nutritional interventions represent a potential non-pharmacological means to increase sleep quality and quantity. (1) Identify an optimal suite of nutritional ingredients and (2) validate the effects of this suite utilising polysomnography, and cognitive and balance tests. The optimal and least optimal combinations of six ingredients were identified utilising 55 male participants and a Box-Behnken predictive model. To validate the model, 18 healthy, male, normal sleepers underwent three trials in a randomised, counterbalanced design: (1) optimal drink, (2) least optimal drink, or (3) placebo were provided before bed in a double-blinded manner. Polysomnography was utilised to measure sleep architecture. Cognitive performance, postural sway, and subjective sleep quality, were assessed 30 min after waking. The optimal drink resulted in a significantly shorter sleep onset latency (9.9 ± 12.3 min) when compared to both the least optimal drink (26.1 ± 37.4 min) and the placebo drink (19.6 ± 32.0 min). No other measures of sleep, cognitive performance, postural sway, and subjective sleep quality were different between trials. A combination of ingredients, optimised to enhance sleep, significantly reduced sleep onset latency. No detrimental effects on sleep architecture, subjective sleep quality or next day performance were observed.