Translating of scientific advances into clinical practice is a major challenge in the stroke research field in the past decades. There were many reasons involved: animal models might not accurately capture all aspects of clinical stroke in humans, the blind and randomized design principle was not closely followed, the inclusion and exclusion criteria was not previously established, sample size was inadequate, endpoint was not scientific nor blindly assessed, inadequate reporting of data and statistical flaws. To bridge the gap between experimental and clinical research, international consortia have attempted to establish standardized guidelines for study design and data report, which include optimizing animal models as well as experimental design, using innovative approaches to assess endpoint, making raw data and negative results available, establishing prior registration mechanism, conducting multicenter preclinical randomized controlled trials (pRCTs), systematic reviews and meta-analysis of preclinical studies, evolving the original focus on neuroprotection into a broader consideration of the role of neurovascular unit and ischemic cascade.