Muscle contractions, among other cellular functions, are driven by highly ordered structures composed of actin and myosin. For cells to properly execute these functions, actin filaments must be specifically organized, assembled, and maintained throughout the cell. The precise timing of actin assembly required for these functions is helped by actin nucleators, such as formins. Formins assemble many actin-based structures through their formin homology (FH) domains, FH1 and FH2. The FH2 domain helps form new filaments in a process called nucleation, while the FH1 domain helps to elongate actin filaments in a processive manner while the FH2 domain “walks” along the growing barbed ends of filaments. The formin homology domain-containing protein (Fhod) family of formins are important for building several contractile actin structures, including sarcomeres in muscle cells and stress fibers in various non-muscle cells. Despite Fhod family formins being required for structures in vivo, mammalian Fhods were initially reported to instead inhibit actin assembly in vitro. Here, we establish that mammalian formin FHOD3 (both isoforms FHOD3S and FHOD3L) nucleate and elongate actin filaments in vitro. Human FHOD3S/L elongate actin filaments quite differently than other formins, where we observe brief, rapid moments of elongation after elongation is paused. We performed rescue experiments for FHOD3L in neonatal rat ventricular myocytes (NRVMs) with mutants that separated its actin assembly activities to better understand whether nucleation or elongation is more important for sarcomere formation. We found that elongation activity by FHOD3L is necessary and sufficient for proper sarcomere formation and maintenance, whereas reducing its nucleation or bundling activity is tolerated in NRVMs. Further, mutations in FHOD3 have been implicated in 1-2% of cases of hypertrophic cardiomyopathy (HCM), a heart disease which results in the thickening of the septal muscle, eventually leading to arrhythmias and heart failure. Interestingly, the likely pathogenic R1386Q mutation for HCM results in a 37% increase in nucleation ability in vitro, leading to thinner sarcomeres in NRVM rescue experiments. We establish NRVMs as a model system to observe sarcomere structure and function and to understand which actin assembly activities are most crucial to the formation and maintenance of the sarcomere.