Background and purpose The onset of wake-up stroke and daytime-unwitnessed stroke is unclear. Though the clinical importance is similar by both being excluded from reperfusion therapy, the characteristics of daytime-unwitnessed stroke are less known than that of wake-up stroke. Here, we compared the characteristics between daytime-unwitnessed stroke and wake-up stroke. Methods Unclear-onset (i.e., last-known normal time ≠ first-found abnormal time) stroke patients admitted within 24 h of recognition of stroke between February 2011 and October 2013 were reviewed. Demographics and clinical and imaging variables were compared between patients with daytime-unwitnessed stroke and those with wake-up stroke. Results Among the 762 ischemic stroke patients, 276 (36.2%) had unclear-onset stroke (104 daytime-unwitnessed stroke and 172 wake-up stroke). Compared to wake-up stroke, daytime-unwitnessed stroke patients had a higher prevalence of cardioembolic stroke and more frequently presented altered mental status ( p < 0.001) and/or aphasia ( p < 0.001) with more severe neurological deficit ( p < 0.001). However, the time from symptom recognition to hospital arrival was shorter ( p < 0.001), and diffusion-weighted image–fluid-attenuated inversion recovery image mismatch ( p = 0.02) and perfusion–diffusion mismatch ( p = 0.001) were also more frequently observed in daytime-unwitnessed stroke. Finally, the proportion of patients eligible for thrombolysis ( p < 0.001) was higher in daytime-unwitnessed stroke patients. Conclusions Clinical and imaging characteristics of daytime-unwitnessed stroke significantly differ from those of wake-up stroke. Daytime-unwitnessed stroke patients are more likely to receive reperfusion therapy, as they arrive at the hospital earlier after symptom recognition, compared to wake-up stroke patients.