Humans live in complex niches where survival and reproduction are conditional on the acquisition of knowledge. The rate at which knowledge is acquired depends on many factors, but its correlation with particular stages of life history might shed some light on the role of childhood in human evolution. Here, we developed a series of Bayesian latent-knowledge models to estimate individual knowledge acquisition, and to infer the influence of age, as well as of activities and social environment on such knowledge. We fit these latent-knowledge models on a set of interviews with 93 children of different ages (4 -26 years of age) from the Island of Pemba, Zanzibar, focusing on their knowledge of the local environment (animals and plants). Compared to previous studies, thanks to the use of latent models, we can describe with more details the variation of knowledge with age as well as other factors, and explore the performance of individuals in different areas of expertise, with question-by-question measurements. In particular, the ability to infer the presence of different dimensions of knowledge, some of which appear to have sex-specific relevance, is a unique feature of our analysis. The results suggest that childhood is a time in which the majority of the knowledge shared between sexes is acquired. Sex differences in total knowledge start to appear during middle childhood, right when labor division emerge. According to the dimension analysis, during adolescence most of the knowledge increases are sex specific. These data and the inferred developmental pattern contribute to a growing comparative literature on children's foraging and the life history of cultural knowledge. ![Kids fishing catfish on Pemba.][1] [1]:https://mfr.de-1.osf.io/export?url=https://osf.io/pr4uw/?direct%26mode=render%26action=download%26public_file=False&initialWidth=675&childId=mfrIframe&parentTitle=OSF+%7C+IMG_0845.JPG&parentUrl=https://osf.io/pr4uw/&format=2400x2400.jpeg