Data comprise mealworm predation rates measured after 24 hours exposure to invertebrates in mature oil palm (2014), and mature and replanted oil palm (2016-2017) plots as part of a large-scale ecological experiment programme (the Biodiversity and Ecosystem Function in Tropical Agriculture project, established in 2013). Eighteen plots were examined across three estates - plots in Ujung Tanjung and Kandista estates were planted in 1987 to 1992 and are mature or over-mature oil palm, while Libo plots (2016-2017 dataset only) were replanted in 2014. Plots were organised in triplets and in in Ujung Tanjung and Kandista, for each triplet one plot was assigned to each of three vegetation treatments: Reduced vegetation cover, normal vegetation management and enhanced vegetation cover. Freshly-killed mealworms (larvae of darkling beetles, Tenebrionidae sp.) were glued onto oil palm fronds trimmed so that ca. 10 cm of each of six leaflets remained. Exclusion and stratum treatments in factorial combinations were applied: caged and uncaged, canopy and ground. The cage exclusion treatments were designed so that most invertebrates could access the fronds but vertebrates could not. Full details about this dataset can be found at https://doi.org/10.5285/03d36ac4-4cf4-46d9-a608-866ba0aab458