This archive contains a unique collection of naturalistic child language data collected between 2017 and 2020 in Southern Senegal. The deposit contains ELAN files of annotated data based on recordings of children's production and child directed speech in naturalistic settings. The language under investigation is Eegimaa, a Jóola language of sourthern Senegal. This is part of the Atlantic branch of the Niger‑Congo Phylum. The data was collected as part of a research project which investigates the acquisition of an Atlantic noun class system. Our research looks at the factors underlying children’s learning of nominal class prefixes and syntactic and semantic agreement at the level of the NP. We focus on questions including the following. • Which elements of noun class morphology do children begin to use productively? • What is the role of input frequency, morphological salience, and transparency in children acquisition of noun class and agreement in Eegimaa? • Are errors in the production of nominal class prefixes also reflected in children’s use of the corresponding agreement markers?