Our purpose was to identify predictors of SE duration and short-term outcome. We performed a prospective study of 95 consecutive patients diagnosed with SE and treated in a neuro-intensive care unit over a period of 3 years. Demographics and clinical data concerning established epilepsy and SE were collected and their relationship to SE duration and short-term outcome was analyzed. The predictive role of non-convulsive SE type, SE treatment with polytherapy, and prior epilepsy with polymorphic seizures for longer SE duration was confirmed on multivariate analysis P < 0.001 (F = 10.89). The longer duration of prior epilepsy and SE proved to be predictors of the rate of recurrent seizures P < 0.001 (F = 14.52). The unfavourable functional outcome correlated with older age, existing neurological abnormalities, mental retardation, prior symptomatic epilepsy, remote symptomatic etiology of SE, non-convulsive SE, and longer duration of SE. Existing neurological abnormalities, mental retardation, SE etiology and duration were confirmed as functional recovery predictors on multivariate analysis P < 0.001 (F = 16.70). The study confirms the predictive value of some clinical factors for SE duration and short-term outcome. Our results are useful for finding more successful strategies in SE management.