Elephant‐shrews (Macroscelidea) have long been considered the only mammalian order to be completely monogamous, based on observations of their pair‐living social organization. We reviewed primary studies on the four components of social systems (social organization, mating system, social structure, and care system) in elephant‐shrews to evaluate whether they truly are monogamous. To identify gaps in our knowledge of their social system, we reviewed evidence for a pair‐living social organization, mate fidelity (mating system), pair bonds (social structure), and biparental care (care system). Field data were available for eight species and seven were often pair‐living. However, these seven species exhibited intra‐specific variation in social organization; two of these species were also solitary living, two species were also group‐living, and the remaining three species were both solitary and group‐living. The eighth species was exclusively solitary. We reconstructed the ancestral social organization of Macroscelidea using Bayesian phylogenetic mixed‐effects models and found that variable social organization, rather than exclusive pair‐living, was the most likely ancestral state, though there was high uncertainty. No socio‐ecological factors (body size, population density, and habitat) predicted a specific social organization. Observations of mating have been rare, such that no firm statements can be made. However, one unpublished study indicated high levels of extra‐pair paternity. Regarding social structure, there was no evidence of pair‐bonding, but there was evidence of mate guarding. Only maternal care has been observed, with females having very short nursing bouts. Evidence suggests that despite having often a pair‐living form of social organization, Macroscelidea should not be described as a monogamous order, as little or no evidence supports that designation, nor are they exclusively pair‐living (social organization) and we urge further field studies on Macroscelidea social systems.