Aims Electromyography (EMG) is a useful tool for assessing the severity of acute peripheral facial nerve palsy (FNP) and predicting the likelihood of recovery. However, the utility of serial EMG studies has rarely been explored, particularly in cases of severe, chronic, or non-idiopathic FNP. The aim of this study was to evaluate the prognostic value of serial EMG in a cohort of patients with severe FNP. Methodology A prospective cohort study was performed on 15 patients with severe FNP presenting to the Sydney Facial Nerve Clinic. Serial EMG studies were performed until facial nerve function stabilised and evaluated denervation potentials, recruitment, and polyphasic action potentials (PAPs). The severity of FNP was assessed using the House-Brackmann (HB) facial grading scale at initial review and most recent follow up, with HB I-IV classified as good recovery. Results The aetiology of FNP included iatrogenic (n=6), Bell’s palsy (n=4), traumatic (n=2), tumour (n=2), and Ramsay Hunt Syndrome (n=1). The median HB score was VI at initial review and IV at follow up, with nine (60%) patients achieving good recovery. EMG was performed once in 15 patients, twice in eight patients, and three times in one patient. The positive predictive value (PPV) for good recovery at initial EMG was 75% for near-normal and normal recruitment and 75% for PAPs. At the second and third EMG, the PPV for good recovery was 100% for no further denervation, 100% for improvement in recruitment, and 100% for PAPs. Conclusion Serial EMG studies enhance the prognostic value of EMG in patients with severe FNP, particularly when the recovery trajectory is uncertain. While serial EMG is effective in identifying patients likely to recover, it may be less reliable for predicting poor outcomes. Nevertheless, clinicians can use serial EMG findings to guide management strategies and provide patients with informed expectations about their recovery.