The presence of cranial autonomic symptoms in migraine is well known and thought to represent activation of the trigeminal parasympathetic reflex pathway similar to trigeminal autonomic cephalalgias. However, studies regarding the prevalence of these symptoms are few. The characteristics of migraineurs with cranial autonomic symptoms and the association of cranial autonomic symptoms with laterality of headache have never been studied in a clinic population. Seventy-eight consecutive subjects with migraine were recruited from the Headache Clinic of the Department of Psychiatry after exclusion of subjects with secondary headache. Their demographic data and detailed history of headache were noted and leading questions were asked regarding cranial autonomic symptoms. χ2 test and Fisher's exact test was used for categorical variables, whereas an independent sample t-test was applied on numerical data. Spearman's correlation was used for correlational analysis of categorical variables. Female subjects (78.2%) outnumbered males and the average duration of illness in the whole sample was 3.81 years. Migraine without aura (53.8%) was the commonest diagnosis, followed by migraine with aura (24.4%). Cranial autonomic symptoms were present in 73.1% of subjects and, commonly, they were ipsilateral to headache. Moreover, strictly unilateral cranial autonomic symptoms were reported by only 32% of patients. The anatomical side of headache did not affect the presence of autonomic symptoms. Those with or without autonomic symptoms did not differ with respect to gender, diagnosis, laterality of headache or associated symptoms except phonophobia, which was more common in subjects with autonomic symptoms (P = 0.05). Those with autonomic symptoms had longer duration of illness (P = 0.03) and longer headache episodes (P = 0.04). In addition, sleep was ineffective in relieving their headache (P = 0.02). Cranial autonomic symptoms are frequent in migraineurs and are common in subjects with long duration of illness and longer headache episodes. Clinical evidence in the present study suggests that subjects with cranial autonomic symptoms have a hyperactive efferent arm of trigeminal autonomic reflex. The connections of trigeminal nucleus with the locus coeruleus and dorsal raphe nucleus may account for the observed phenotypic differences between the two groups. Further research, however, is required to elucidate the underlying neural mechanisms of cranial autonomic symptoms in migraine.