Weapon carrying has detrimental effects for perpetrators and victims alike. It is therefore imperative that research efforts are invested into establishing those contextual factors that are associated with this antisocial behavior. This systematic and meta-analytic review investigates the association of weapon carrying with bullying perpetration and victimization inside and outside the school context. Results on pure bullies, pure victims and bully-victims are also presented.

Further to extensive searches, across 20 databases and 46 journals, and careful screening of reports, in line with pre-established methodological criteria, a total of 35 manuscripts are included in the meta-analysis. Narrative results based on longitudinal studies are also presented but not meta-analyzed given the variability in study characteristics and the small number of studies.

Weapon carrying is significantly associated with both bullying perpetration (adjusted $OR$ = 2.64; $p$ < 0.001) and victimization (adjusted $OR$ = 1.58; $p$ < 0.05). Effect sizes are larger when looking at discrete categories of pure bullies (adjusted $OR$ = 3.24; $p$ < 0.01), pure victims (adjusted $OR$ = 1.79; $p$ < 0.05) and bully-victims (adjusted $OR$ = 5.66; $p$ < 0.001) when compared with non-involved school children. Subgroup analyses suggest that pure victims ($Q$ = 6.77; $p$ < 0.01) and bully-victims ($Q$ = 8.01; $p$ < 0.01) are significantly more likely to carry a weapon inside than outside the school, thus rendering support to the ‘vulnerability/self-protection’ hypothesis. Pure bullies have the same odds of carrying a weapon inside and outside the school context ($Q$ = 0.60; $p$ = 0.44), supporting a persistent antisocial personality theoretical framework.

Implications for policy and practice arising from our results are discussed.