To identify clinical and predictive features of outcome in cryptogenic epilepsy in pediatric neurology practice, the medical records of all patients with cryptogenic epilepsy (as defined by the International League Against Epilepsy) in a single pediatric neurology practice over a 12-year interval with at least 2 years of follow-up were systematically and retrospectively reviewed. Review revealed 60 children with cryptogenic epilepsy: 32 (53.3%) males, 11 (18.3%) prior febrile seizure, 9 (15.0%) developmental delay at onset, and 38 (63.3%) placement in regular classes. Twenty-two (35.7%) had generalized seizures. Mean follow-up after initiating antiepileptic medication was 53 months (range 24-128 months). Four (6.7%) were intractable; 4 (6.7%) had very poor outcomes; 8 (13.3%) had poor outcomes; 44 (73.3%) were well controlled. Sixteen (26.7%) and 31 (51.7%) had seizure recurrence within the last 12 and 24 months, respectively. Twenty-nine (48.3%) were seizure-free for at least 24 months. Factors associated with a poor outcome include seizure recurrence in the 6- to 12-month interval after therapy initiation (P = 0.006) and developmental delay at onset (P = 0.023). This case series suggests that children with cryptogenic epilepsy tend to have a favorable outcome. Seizure recurrence in the first months after therapy initiation and developmental delay apparent at onset are predictive of poor outcome.