Mutations in the co-chaperone Bcl2-associated athanogene 3 (BAG3) can cause myofibrillar myopathy (MFM), a childhood-onset progressive muscle disease, characterized by the formation of protein aggregates and myofibrillar disintegration. In contrast to other MFM-causing proteins, BAG3 has no direct structural role, but regulates autophagy and the degradation of misfolded proteins. To investigate the mechanism of disease in BAG3-related MFM, we expressed wild-type BAG3 or the dominant MFM-causing BAG3 (BAG3^P209L) in zebrafish. Expression of the mutant protein results in the formation of aggregates that contain wild-type BAG3. Through the stimulation and inhibition of autophagy, we tested the prevailing hypothesis that impaired autophagic function is responsible for the formation of protein aggregates. Contrary to the existing theory, our studies reveal that inhibition of autophagy is not sufficient to induce protein aggregation. Expression of the mutant protein, however, did not induce myofibrillar disintegration and we therefore examined the effect of knocking down Bag3 function. Loss of Bag3 resulted in myofibrillar disintegration, but not in the formation of protein aggregates. Remarkably, BAG3^P209L is able to rescue the myofibrillar disintegration phenotype, further demonstrating that its function is not impaired. Together, our knockdown and overexpression experiments identify a mechanism whereby BAG3^P209L aggregates form, gradually reducing the pool of available BAG3, which eventually results in BAG3 insufficiency and myofibrillar disintegration. This mechanism is consistent with the childhood onset and progressive nature of MFM and suggests that reducing aggregation through enhanced degradation or inhibition of nucleation would be an effective therapy for this disease.