We examined whether French children in Grades 3 and 5 (aged ∼ 8–11 years) benefit from morphological relatedness beyond orthographic relatedness in the implicit learning of new spellings. Children silently read stories that included two target nonwords. One nonword was in an opaque condition in that nothing in the story could justify the spelling of its final sound. The other nonword was in either a morphological condition (for children in the morphological group) or an orthographic condition (for children in the orthographic group). In the morphological condition, the final spelling of the target nonword was justified by two morphologically related nonwords. For example, coirardage, obtained by adding the suffix age to coirard, designates the coirard’s song and justifies the final silent d of coirard. The orthographic condition included two nonwords that were orthographically but not morphologically related to the target. For example, the coirard’s song was coirardume, obtained by adding ume, which is not a suffix, to coirard. Then, 30 min after reading the stories, children were asked to choose the correct spelling of each nonword from among three phonologically plausible alternatives (e.g., coirard, coirars, coirar). In the morphological group, both third and fifth graders more often selected the correct spellings for items presented in the morphological condition than for items presented in the opaque condition. In the orthographic group, the results were very similar in the opaque and orthographic conditions. The findings show that the benefit of morphological relatedness in the implicit learning of new spellings cannot be reduced to orthographic relatedness.< Réduire