To determine clinical predictors contributing to the diagnosis of children with a neurologic catastrophe who present with headache. A 10-year retrospective case-control study of all children, aged 4–18, with no known neurologic disorder, diagnosed with a neurologic catastrophe (brain tumour, cerebral haemorrhage, pseudotumor cerebri, cerebro-vascular malformation, brain abscess and acute hydrocephalus) who presented to our centre with a chief complaint of headache. The characteristics of the identified cases and there clinical manifestations were compared with matched controls complaining of headaches found to have a normal brain imaging. A logistic regression included all the variables found to have a significant association in a univariate analysis (Mann-Whitney U and Chi-square tests). Thirty cases (4 patients with astrocytomas, 6 pseudotumors cere-bri, 2 glioblastomas, 9 cerebral haemorrhage, 2 ependymomas, 3 medulloblas-tomas, 2 brain abscesses, 1 pineal dysgerminoma, 1 choroid plexus carcinoma, 1 optic glioma) and 30 controls were identified and compared. The majority of children suffering from a neurologic catastrophe had an abnormal neurological exam (p<0.0001), had no family history of migraine (p<0.006), presented vomiting (p<0.008) and had the headache for a shorter duration 5 weeks ±90 (SD) neurologic catastrophe VS 71 weeks ±108 (SD) other headache; p<0.001). On history, not having a family history of migraine resulted in an odds ratio of 12 (p<0.05) and having a history of vomiting resulted in an odds ratio of 17.5 (p<0.05). Having an abnormal neurological exam was associated with an odds ratio of 278 (p<0.05). We consider that our clinical predictors could help in guiding the appropriate use of brain imaging when they will be tested in a prospective study.