In a previous trial in Mali, we showed that traditional pearl millet couscous and thick porridge delayed gastric emptying (about 5 h half-emptying times) in a normal-weight population compared with non-traditional carbohydrate-based foods (pasta, potatoes, white rice; about 3 h half-emptying times), and in a gastric simulator we showed millet couscous had slower digestion than wheat couscous. In light of these findings, we tested the hypothesis in a normal-weight US population (n 14) that millet foods would reduce glycaemic response (continuous glucose monitor), improve appetitive sensations (visual analogue scale ratings), as well as reduce gastric emptying rate (13C-octanoic acid breath test). Five carbohydrate-based foods (millet couscous – commercial and self-made, millet thick porridge, wheat couscous, white rice) were fed in a crossover trial matched on available carbohydrate basis. Significantly lower overall glycaemic response was observed for all millet-based foods and wheat couscous compared with white rice (P ≤ 0·05). Millet couscous (self-made) had significantly higher glycaemic response than millet couscous (commercial) and wheat couscous (P < 0·0001), but as there were no differences in peak glucose values an extended glycaemic response was indicated for self-made couscous. Millet couscous (self-made) had significantly lower hunger ratings and higher fullness ratings (P < 0·05) than white rice, millet thick porridge and millet couscous (commercial). A normal gastric emptying rate (<3 h half-emptying times) was observed for all foods, with no significant differences among them. In conclusion, some traditionally prepared pearl millet foods show the potential to reduce glycaemic response and promote satiety.