Background: Subarachnoid hemorrhage (SAH) is a devastating neurological disease associated with high rates of mortality and disability. Aneurysms are the main cause of non-traumatic subarachnoid hemorrhages. However, non-traumatic non-aneurysmal subarachnoid hemorrhage (naSAH), another clinical type of SAH, has been poorly studied for its prognosis and risk factors. Method and result: We collected demographic and clinical variables for 126 naSAH and 89 aneurysmal subarachnoid hemorrhage (aSAH) patients, including age and gender; hospitalization days; hematological indicators; clinical score scales; past medical history; and personal history. We found that the monocytes in naSAH (0.50 ± 0.26) patients were lower than in aSAH patients (0.60 ± 0.27). The prevalence of diabetes in naSAH (30.2%) patients was higher than in aSAH (14.5%) patients. The naSAH patients were divided into good and poor outcome groups based on the modified Rankin Scale at the 90th day (90-day mRS) after discharge. A univariate analysis showed that there were significant differences in age, white blood cell count (WBC), monocyte count, D-dipolymer, neuron-specific enolase (NSE), random blood glucose (RBG), aspartate transaminase (AST), urea and free triiodothyronine (FT3) between the two groups. A logistic regression showed that aging and high level NSE were independent risk factors for a poor outcome. The predictive ability of age (area under curve (AUC) = 0.71) and NSE (AUC = 0.68) were analyzed by a receiver operating characteristic (ROC) curve. The results of the logistic regression suggested that age, D-dipolymer, NSE, RBG, urea and FT3 distinguished and predicted the prognosis of naSAH. The discriminant analysis of the above variables revealed that the discriminant accuracy was 80.20%. Conclusions: Compared with aSAHs, naSAHs are more likely to occur in patients with diabetes, and the level of monocytes is lower. Moreover, the prognosis of elderly patients with an naSAH is relatively poor, and the level of NSE in the course of the disease also reflects the prognosis. Multivariate comprehensive analysis is helpful to judge the prognosis of patients at a small cost.