The effect of the babyface schema includes three typical responses, namely, the preference response, viewing motivation, and attention bias towards infant faces. It has been theorised that these responses are primarily influenced by infants’ facial structures. However, recent studies have revealed the moderating role of facial expression, suggesting that the strongest effect of the babyface schema may be related to the neutral facial expression; this hypothesis remains to be tested. In this study, the moderating role of facial expression was assessed in three successive experiments (total N = 402). We used a series of images of the same face with multiple expression-standardised images of infants and adults to control for facial structure. The results indicated that the effect sizes of the babyface schema (i.e., response differences between infants and adults) were different for multiple expressions of the same face. Specifically, the effect sizes of neutral faces were significantly greater than those of happy and sad faces according to the preference response (experiment 1, N = 90), viewing motivation (experiment 2, N = 214), and attentional bias (experiment 3, N = 98). These results empirically confirm that neutral infant facial expressions elicit the strongest effect of the babyface schema under the condition of using adult faces as a comparison baseline and matching multiple expressions of the same face.