The role of hedonic reactions to odours on determining rejections of foods in infants is still debated. To answer this question, the first requirement is to have an instrument to test infants’ olfactory reactivity. Indeed, except in neonates, there is a lack of validated method. This study presents results of a new method designed to assess olfactory reactivity of 7- to 8-month-old infants. This method is based on the observation of infants’ reactions while they explored different rattles. The rattles were visually identical but differed in their odours (neutral or with a food odour). Two sequences of 6 rattles were presented to each infant. Eight odorants were used. Four odorants were chosen to evoke odours of often rejected foods and four were chosen to evoke odours of often liked foods. The hedonic valence was previously checked with an adult panel. These odours were also equated for intensity and had a very low perceived trigeminal component. Forty infants were tested and videotaped. These videotapes were subjected to frame-by-frame analysis to measure a variety of behaviours supposed to reflect the infants’ hedonic reaction (e.g., times spent to mouthing or pushing away the rattle). Duration of these events for each odorized rattle was compared to those observed for the control rattles. It thus allows getting a reactivity index per odorant and over all odorants. Reactivity of infants to these odorants was related to infants’ reactions to food during weaning, evaluated by the mother. Methodological considerations concerning the reliability of measures of hedonic reactivity in infants will be discussed. The issue of whether young children experience odours as pleasant or unpleasant will be discussed in relation with their food experiences.