Background: Serum prolactin level has been used as an alternative determinant to help differentiate epileptic from non-epileptic seizures. We aimed to analyze the association between serum prolactin concentration and epileptic seizure versus febrile seizure as well as seizure duration, single versus multiple seizure attacks and time lapse between seizure onset and blood sampling. Methods: Seventy patients aged 6 months to 5 years with seizure who admitted to the emergency department between March 2004 and February 2005 were selected and divided into group I (epileptic seizure) and group II (febrile seizure). Group III consisted of 35 pediatric patients without seizure. Blood samples were collected within 2 hours after seizure. Serum prolactin levels were measured, and statistical analyses were performed to detect possible associations between variables. Results: Serum prolactin level increased about twice the normal level after epileptic seizures but not after febrile seizures or in the control group. The length of seizure and multiple seizure attacks were positively associated, whereas the lapse between seizure onset and blood drawing was negatively correlated with an increase in serum prolactin level. Conclusion: Febrile (non-epileptic) seizures were not associated with increased serum prolactin level. Elevated prolactin level within 2 hours of a seizure may be suggestive of epileptic origin of the seizure.