Data on the acute natural history of progressive stroke with or without heparin treatment are limited. To define the acute course of patients treated with heparin for progressive stroke, we examined the charts of 69 such patients identified through the Cornell Neurology Database from October 1979 to June 1985. Analysis included determining whether further clinical deterioration or hemorrhagic complications were associated with readily identifiable clinical or laboratory variables. Twenty-five patients (36%) continued to deteriorate while receiving heparin, another two (3%) worsened due to intracerebral hemorrhage, and a total of 10 patients (14%) had bleeding complications. No clinical features or heparin dosing regimens distinguished the patients likely to benefit from heparin. Clinical progression or hemorrhage did not correlate with the level of anticoagulation as measured by the average heparin dose per day or the mean partial thromboplastin time. Without unequivocal evidence demonstrating heparin's ineffectiveness for progressive stroke, many clinicians managing such patients will continue to use heparin. Our results suggest that this decision should not be governed by such clinical features as a patient's age or sex or by the vascular distribution of the stroke. Furthermore, frequent measurement of and overzealous efforts to adjust the partial thromboplastin time may be unnecessary since it does not correlate with outcome.