Background and Purpose-Less than 25% of stroke patients arrive to an emergency department within the 3-hour treatment window. Stroke Warning Information and Faster Treatment (SWIFT) compared an interactive intervention (II) with enhanced educational (EE) materials on recurrent stroke arrival times in a prospective cohort of multiethnic stroke/ transient ischemic attack survivors. Methods-A single-center randomized controlled trial (2005)(2006)(2007)(2008)(2009)(2010)(2011) randomized participants to EE (bilingual stroke preparedness materials) or II (EE plus in-hospital sessions). We assessed differences by randomization in the proportion arriving to emergency department <3 hours, prepost intervention arrival <3 hours, incidence rate ratio for total events, and stroke knowledge and preparedness capacity. Results-SWIFT randomized 1193 participants (592 EE, 601 II): mean age 63 years; 50% female, 17% black, 51%Hispanic, 26% white. At baseline, 28% arrived to emergency department <3 hours. Over 5 years, first recurrent stroke (n=133), transient ischemic attacks (n=54), or stroke mimics (n=37) were documented in 224 participants. Incidence rate ratio=1.31 (95% confidence interval=1.05-1.63; II to EE). Among II, 40% arrived <3 hours versus 46% EE (P=0.33). In prepost analysis, there was a 49% increase in the proportion arriving <3 hours (P=0.001), greatest among Hispanics (63%, P<0.003). II had greater stroke knowledge at 1 month (odds ratio=1.63; 1.23-2.15). II had higher preparedness capacity at 1 month (odds ratio=3.36; 1.86, 6.10) and 12 months (odds ratio=7.64; 2.49, 23.49). Conclusions-There was no difference in arrival <3 hours overall between II and EE; the proportion arriving <3 hours increased in both groups and in race-ethnic minorities. Clinical Trial Registration-URL: http://www.clinicaltrials.gov. Unique identifier: NCT00415389.