Background:: In some cases of spontaneous subarachnoid hemorrhage (SAH), the cause of bleed remains obscure on initial evaluation. These patients may harbor structural lesions. We aim to determine the utility of repeat angiogram in these subsets of patients. Methods:: In this prospective study, patients with SAH with a negative computed tomographic angiogram (CTA) and digital subtraction angiogram (DSA) were included. A repeat angiogram was done after 6 weeks of initial angiogram. Patients were divided into perimesencephalic SAH (PM-SAH) and diffuse classic SAH (Classic-SAH) groups. Outcome was determined by modified Rankin score (mRS). Results:: A total of 22% (39/178) of all SAH were angio-negative. A total of 90% (n = 35) of these were in Hunt and Hess grade 1-3. A total of 22 patients had PM-SAH and 17 had a Classic-SAH. Repeat angiogram did not reveal any pathology in the PM-SAH group, whereas two patients with Classic-SAH were found to have aneurysms. At 6 months follow-up, 95% patients of PM-SAH and 83.3% of Classic-SAH had mRS of 0. Conclusion:: Repeat angiogram is probably not necessary in patients of PM-SAH and they tend to have better outcome. Classic-SAH pattern of bleed is associated with fair chances of an underlying pathology and a repeat angiogram is recommended and these cases and they have poorer outcome.