Animals, including insects, need oxygen for aerobic respiration and eventually asphyxiate without it. Aerobic respiration, however, produces reactive oxygen species (ROS), which contribute to dysfunction and aging. Animals appear to balance risks of asphyxiation and ROS by regulating internal oxygen relatively low and stable, but sufficient levels. How much do levels vary among species, and how does variation depend on environment and life history? We predicted that lower internal oxygen levels occur in insects with either limited access to environmental oxygen (i.e., insects dependent on aquatic respiration, where low internal levels facilitate diffusive oxygen uptake, and reduce asphyxiation risks) or consistently low metabolic rates (i.e., inactive insects, requiring limited internal oxygen stores). Alternatively, we predicted insects with long life-stage durations would have internal oxygen levels > 1 kPa (preventing high ROS levels that are believed to occur under tissue hypoxia). We tested these predictions by measuring partial pressures of oxygen (PO2) in tissues from juvenile and adult stages across 15 species comprising nine insect orders. Tissue PO2 varied greatly (from 0 to 18.8 kPa) and variation across species and life stages was significantly related to differences in habitat, activity level, and life stage duration. Individuals with aquatic respiration sustained remarkably low PO2 (mean = 0.88 kPa) across all species from Ephemeroptera (mayflies), Plecoptera (stoneflies), Trichoptera (caddisflies), and Diptera (true flies), possibly reflecting a widespread, but hitherto unknown, adaptation for extracting sufficient oxygen from water. For Odonata (dragonflies), aquatic juveniles had higher PO2 levels (mean = 6.12 kPa), but these were still lower compared to terrestrial adults (mean = 13.3 kPa). Follow-up tests in juvenile stoneflies showed that tissue PO2 remained low even when exposed to hyperoxia, suggesting that levels were down-regulated. This was further corroborated since levels could be modulated by ambient oxygen levels in dead individuals. In addition, tissue PO2 was positively related to activity levels of insect life stages across all species and was highest in stages with short durations. Combined, our results support the idea that internal PO2 is an evolutionarily labile trait that reflects the balance between oxygen supply and demand within the context of the environment and life-history of an insect.