To determine whether differences exist between general emergency physicians (GEMs) and pediatric emergency physicians (PEMs) in the emergency care of children with common pediatric emergencies. We carried out a survey study of all members of the American Academy of Pediatrics Section of Emergency Medicine and the Washington State American College of Emergency Physicians. We identified current therapeutic interventions for croup, asthma, bronchiolitis, seizures, febrile infant, conscious sedation, head trauma, and coin ingestion, and compared the practice patterns of GEMs and PEMs. A total of 66% of the surveys were returned, including 211 GEMs and 329 PEMs. The majority of PEMs practice in children's hospitals, whereas most GEMs practice in general community hospitals. Slightly over half (51%) of PEMs are PEM fellowship-trained versus 1% of GEMs. CROUP: The majority of GEMs and PEMs use racemic epinephrine (RE) in the treatment of a child with stridor at rest; approximately one-third admit to the hospital after RE (39 vs 30%, NS). PEMs are more likely to observe the child for >2 hours after RE (94% vs 79%, P < 0.01). The majority of PEMs and GEMs use steroids in these patients (94 vs 88%, NS). ASTHMA: There is no significant difference in the use of albuterol, aminophylline, or steroids. Steroids are more likely to be given orally by PEMs than GEMs (74 vs 50%, P < 0.01). BRONCHIOLITIS: The majority of both groups of physicians routinely use nebulized beta-agonists; however, significantly more GEMs than PEMs use steroids (68 vs 45 %, P < 0.01). Half of GEMs vs 78% of PEMs use lorazepam as a first line drug in the treatment of seizures (P < 0.01). There is no significant difference with respect to the use of rectal diazepam in the pre-hospital setting. FEBRILE INFANT: GEMs are less likely than PEMs to admit the febrile infant <4 weeks of age (68 vs 87%; P < 0.01). Admission of older febrile infants (four to six weeks and eight weeks of age) is not significantly different between PEMs and GEMs. CONSCIOUS SEDATION: Both groups use a wide array of drugs alone or in combination to sedate children for complex facial laceration repair, closed fracture reduction, and cranial computed tomography (CT). GEMs are more likely to use ketamine for laceration repair (28 vs 16%, P < 0.01). Both GEMs and PEMs use midazolam plus a narcotic for fracture reduction. For further sedation for cranial CT, after an initial dose of midazolam, GEMs are more likely to use additional midazolam (64 vs 47%, P < 0.01), and PEMs are more likely to add pentobarbital (15 vs 4%, P < 0.01). HEAD TRAUMA: Most GEMs (87%) and PEMs (81%) would obtain a cranial CT on a neurologically normal two year old who had fallen down the stairs with a six-minute loss of consciousness. COIN INGESTION: Most GEMs and PEMs would obtain radiographs on an asymptomatic two year old with a recent coin ingestion. With some notable exceptions, GEMs and PEMs have similar pediatric practice patterns despite differences in training and practice environments.