Fibrodysplasia ossificans progressiva (FOP) is a rare genetic disorder characterized by episodic but cumulative heterotopic ossification (HO). The HO in FOP develops by an endochondral process that involves a cartilage intermediate and which overall appears to mirror normal endochondral bone formation. As this heterotopic bone develops, it connects to the normal skeleton (by a process that is currently not understood), and these connections eventually start to bridge joints, resulting in progressive immobility, and eventually death. FOP is caused by mutations in the type I BMP (bone morphogenetic protein) receptor ACVR1, the most mutation altering arginine 206 to histidine (ACVR1R206H). We have generated a genetically and physiologically accurate mouse model of FOP – 129;B6N-Acvr1tm2Vlcg (Acvr1[R206H]FlEx)/+ Gt(ROSA26)Sortm3.1(cre/ERT2)Vlcg/+. In this model, the HO process is initiated by treating the mice with tamoxifen, that effectively converts the conditional allele – Acvr1[R206H]FlEx – to its mutant counterpart – Acvr1R206H – thereby generating mouse cohorts with the same genotype as FOP patients. Using this model, we demonstrated that there is an absolute requirement for activin A to kickstart the process of HO. This requirement for activin A was demonstrated in vivo using neutralizing antibodies to activin A. However, in these experiments the anti-activin A antibody was administered at the same time as initiation of HO, and hence this setting only addressed the requirement of activin A at the initial stages of HO, prior to the formation of cartilage. We explored whether activin A continues to play a role after the endochondral process has started and even when partly mineralized heterotopic lesions could be observed. Our results indicate that inhibition of activin A with a fully human monoclonal antibody completely blocks formation of HO in this model of FOP when the antibody is administered 9 days after initiation of the model. To further test delayed treatment’s effect on already formed – yet still nascent – HO, anti-activin A Ab was injected at 3 weeks post initiation of the model. With 3-weeks treatment, existing HO volume (lesion size) decreased by 37±41 mm3 in activin A Ab treatment group (p<0.05), while it increased by 19±45 mm3 in isotope control group. In summary, these results demonstrate a continued role for activin A even after mineralized heterotopic bone lesions have formed and indicates that anti-activin A may be able to reverse the progression of nascent (and not fully mineralized) heterotopic bone.