The relationship between emotion and language in children is an emerging field of research. To carry out this type of study, researchers need to precisely manipulate the emotional parameters of the words in their experimental material. However, the number of affective norms for words in this population is still limited. To fill this gap, the present report presents a set of 7 norms that measure the emotional parameters (valence, arousal) of words rated by children in French, English, German, Spanish, and Chinese. The high correlations between the valence values and the moderate correlations between arousal values of these norms allow us to determine a cross-language homogeneity of emotional representations in children. Thus, this report introduces a significant metanorm with 508 words characterized by valence and by arousal. This tool is a worthwhile resource for researchers interested in the links between emotion and language, as well as for those interested in cross-language comparisons. The complete database can be downloaded at the following address: https://osf.io/stnbk/