The growing disconnect between children and nature has led to concerns around the attrition of ecological knowledge, reduced opportunities for developing a connection with nature at a young age and the associated potential of waning support for future conservation. Understanding perceptions of the natural world among children is vital to engage them with local wildlife and mitigate the growing disconnect between children and nature. This study investigated children’s perceptions of nature by analysing drawings made by children of their local green spaces.

We collected 401 drawings by children aged 7–11 years, from 12 different primary schools (including state-funded and privately funded schools) in England, and assessed which animal and plant groups were drawn the most and least often. We quantified the species richness and community composition of each drawing and identified all terms used in the labels and captions to the highest level of taxonomic resolution possible. We estimated the greenness of each school’s surroundings and tested whether there were differences in taxonomic resolution between taxa and whether school type or school greenness affected animal or plant species richness, community composition or taxonomic resolution of terms.

Mammals and birds were the most commonly drawn groups, appearing in 80.5% and 68.6% of drawings respectively. The least commonly drawn group was herpetofauna, appearing in just 15.7% of drawings. Despite not explicitly being asked about plants, 91.3% of drawings contained a plant.

Taxonomic resolution was higher for mammals and birds than other taxa, with 90% of domestic mammals and 69.6% of garden birds identifiable to species, compared to 18.5% of insects and 14.3% of herpetofauna. No invertebrates other than insects were identifiable to species. Within plants, trees and crops were the most identifiable to species, at 52.6% and 25% of terms respectively.

There was no difference in animal species richness or plant community composition between school types, but drawings from state-school children had higher plant species richness than those from private-school children. Animal community composition was different between school types, with more types of garden birds drawn by private-school children than state-school children, and more types of invertebrates drawn by state-school children than private-school children. School greenness had no effect on species richness or on plant community composition, but we found an effect on animal community composition. There was no difference in the taxonomic resolution of terms between school types.

Our findings indicate that children’s awareness of local wildlife is focused on mammals and birds, with significantly less attention given to invertebrates and herpetofauna. While plants feature highly in children’s perceptions of nature, their knowledge is less specific than that of animals. We suggest that this skew in children’s ecological knowledge towards familiar, charismatic animals be addressed through better integration of ecology within national curricula and more funding for green space within school grounds since smaller invertebrate taxa are likely to be more abundant in these spaces. Encouragingly, levels of awareness did not systematically differ between school types, indicating that similar approaches to engage children with nature should be made across school types.