Objectives Histomorphological analyses of bones are used to estimate an individual's chronological age, interpret a bone's load history, and differentiate species. Among various histomorphological characteristics that can influence mechanical properties of cortical bone, secondary osteon (Haversian system) population density and predominant collagen fiber orientation are particularly important. Cross‐sectional shape characteristics of secondary osteons (On.Cr = osteon circularity, On.El = osteon ellipticality) are considered helpful in these contexts, but more robust proof is needed. We sought to determine if variations in osteon shape characteristics are sufficient for accurately differentiating species, load‐complexity categories, and regional habitual strain‐mode distributions (e.g., tension vs. compression regions). Materials and Methods Circularly polarized light images were obtained from 100‐micron transverse sections from diaphyses of adult deer calcanei; sheep calcanei, radii, and tibiae; equine calcanei, radii, and third metacarpals (MC3s); chimpanzee femora; and human femora and fibulae. Osteon cross‐sectional area (On.Ar), On.Cr, and On.El were quantified indiscriminately and in the contexts of load‐complexity and regional strain‐mode distributions. Results On.Cr and On.El, when examined independently in terms of all data, or mean (nested) data, for each bone, exceeded 80% accuracy in the inter‐species comparisons only with respect to distinguishing humans from nonhumans. Correct classification among the nonhuman species was <70%. When On.Cr and On.El were coupled together and with On.Ar in discriminant function analyses (nested and unnested data) there were high misclassifications in all but human vs. nonhuman comparisons. Discussion Frequent misclassifications in nonhuman comparisons might reflect influences of habitual load complexity and/or strain‐mode distributions, or other factors not accounted for by these two considerations.