Introduction: Caregiver burden is well characterized in adults with psychosis but less is known about the burden experienced by carers of children with psychosis. In this secondary analysis, we examined burden reported by caregivers of children with clinical high risk for psychosis and psychotic disorders to test an explanatory model for caregiver burden. Methods: Cross-sectional analyses examined burden experienced by caregivers of children with clinical high risk for psychosis and psychosis who were recruited from an outpatient psychiatric clinic. Caregiver burden was assessed with the Zarit Burden Inventory. Structural equation modelling was employed to examine a model assessing the simultaneous effects of factors related to burden among caregivers of children at clinical high risk for psychosis and diagnosed with psychotic disorders. Results: The study involved 85 caregivers of children aged 6-17. The latent variable for positive caregiving attributes, which included parenting confidence and communication, had a direct effect on caregiver burden, with greater levels of positive caregiving attributes associated with decreased burden. Direct effects of social support and the latent variable for primary/secondary stressors on burden were not observed. There was a significant indirect effect for social support through positive caregiving attributes to mitigate the effects of primary/secondary stressors on caregiver burden. Conclusion: The results provide preliminary evidence supporting the importance of social support, parenting confidence, and caregiver-child communication as key targets for intervention. Moreover, our findings suggest that despite stressors, such as suicidal behaviour and psychiatric hospitalisations, modifiable factors exist to serve as protective buffers against caregiver burden.