To date, there have been no population-based data published regarding the influence of the patient's demographic factors on rt-PA use. We present preliminary data regarding the effect of race and gender on patterns of rt-PA use in the pre-FDA approval era, in a population with demographic and socioeconomic characteristics similar to the United States. All ischemic strokes within a biracial population of 1.3 million were identified by review of all primary and secondary hospital ICD-9-CM codes 430-438 from July 1993 to June 1994 at all hospitals in the region. The number of patients treated with rt-PA or placebo as part of the blinded NINDS rt-PA trial, as well as demographic characteristics, were recorded and analyzed. There were a total of 1973 hospitalized ischemic strokes that occurred at a hospital participating in the NINDS rt-PA trial. Patients that received rt-PA were significantly younger than those that did not (mean age 67 v 72, respectively, P = .01). Of the 413 strokes that occurred in African Americans, 2.2% were treated with rt-PA vs. 2.6% of the 1560 non-African Americans. Women (2.0%) and men (3.0%) were equally likely to receive rt-PA. The single academic center was as likely to give rt-PA as the community medical centers. In the Greater Cincinnati/Northern Kentucky population, patterns of rt-PA use in 1993-94 did not appear to vary according to race or gender, or type of medical center. These findings may be in part because of the regionally-based method of stroke care delivery in the area.