Recently, recessive mutations of MAGI2 were identified as a cause of steroid-resistant nephrotic syndrome (SRNS) in humans and mice. To further delineate the pathogenesis of MAGI2 loss of function, we generated stable knockout lines for the two zebrafish orthologues magi2a and magi2b by CRISPR/Cas9. We also developed a novel assay for the direct detection of proteinuria in zebrafish independent of transgenic background. Whereas knockout of magi2b did not yield a nephrotic syndrome phenotype, magi2a -/larvae developed ascites, periorbital edema, and proteinuria, as indicated by increased excretion of low molecular weight protein. Electron microscopy demonstrated extensive podocyte foot process effacement. As in human SRNS, we observed genotype/phenotype correlation, with edema onset occurring earlier in zebrafish with truncating alleles (5-6 days post fertilization) versus hypomorphic alleles (19-20 days post fertilization). Paradoxically, corticosteroid treatment exacerbated the phenotype, with earlier onset of edema. In contrast, treatment with cyclosporine A or tacrolimus had no significant effect. Although RhoA signaling has been implicated as a downstream mediator of MAGI2 activity, targeting of the RhoA pathway did not modify the nephrotic syndrome phenotype. In the first CRISPR/Cas9 zebrafish knockout model of SRNS, we found that corticosteroids may have a paradoxical effect in the setting of specific genetic mutations.