Objective: To determine whether onset seizures after subarachnoid hemorrhage (SAH) carry independent prognostic information and to investigate the risk factors for late seizures after SAH. Background: Modern management of SAH, including early operation, has substantially reduced mortality. No study has adequately assessed the importance of onset seizures in a contemporary SAH cohort. Methods: The authors analyzed the records and initial CT scans of 412 consecutive patients with aneurysmal or nonaneurysmal SAH admitted to the Royal Melbourne Hospital from 1990 to 1996. Each patient with an onset seizure (n = 32, 7.8% of cohort) was age and sex matched to two nonseizure patients of the same cohort. Each patient with a late seizure (n = 17, 5.1% of cohort) was matched to five control subjects of the same cohort. Results: With use of logistic regression analysis, onset seizures correlated with the sum score of blood on initial CT scan (OR = 1.1, p = 0.05), but there was no significant correlation with duration of loss of consciousness at onset, Glasgow Coma Score (GCS), presence of aneurysm, or past history of hypertension or epilepsy. Disability 6 weeks after SAH according to the Glasgow Outcome Scale was independently predicted by initial GCS of <6 (OR = 13.7, p < 0.01) and onset seizure (OR = 7.8, p = 0.04). Late seizures within the first 6 weeks were independently related to rebleeding (OR = 94, p < 0.01) and onset seizures (OR = 27, p < 0.01) but not to other onset variables, development of hydrocephalus, or vasospasm. Conclusion: In this single-institution cohort of patients with SAH, onset seizures were an independent risk factor for late seizures and a predictor of poor outcome.