A variety of methods was used to compare patient mix, practice variation, survival, and recurrence after first ischemic stroke among Rochester, MN residents. The significance of the results for neurologists and generalists was examined. Age, stroke severity, congestive heart failure (CHF), and the interaction between atrial fibrillation and patient groups were determinants of survival. Without atrial fibrillation, patients on neurology services and patients on general services with neurology consultation had better survival than those without neurology consultation, adjusting for age, stroke severity, and CHF. With atrial fibrillation, patients on general services with neurology consultation had no better survival compared with those without neurology consultation; patients on neurology services had worse survival (p=0.002). There was no difference in stroke recurrence. Evaluation by a neurologist is associated with better survival for most patients with ischemic stroke but not those with atrial fibrillation. Only a randomized trial can determine whether this association is causal.