Hydrogen sulfide (H2S) is a gasotransmitter that has been studied for its potential therapeutic effects, including its role in the pathophysiology and treatment of stroke. This systematic review and meta-analysis aimed to determine the sufficiency of overall preclinical evidence to guide the initiation of clinical stroke trials with H2S and provide tailored recommendations for their design. PubMed, Web of Science, Scopus, EMBASE, and MEDLINE were searched for studies evaluating the effect of any H2S donor on in vivo animal models of regional ischemic stroke, and 34 publications were identified. Pooling of the effect sizes using the random-effect model revealed that H2S decreased the infarct area by 34.5% (95% confidence interval (CI) 28.2–40.8%, p < 0.0001), with substantial variability among the studies (I2 = 89.8%). H2S also caused a 37.9% reduction in the neurological deficit score (95% CI 29.0–46.8%, p < 0.0001, I2 = 63.8%) and in the brain water content (3.2%, 95% CI 1.4–4.9%, p = 0.0014, I2 = 94.6%). Overall, the studies had a high risk of bias and low quality of evidence (median quality score 5/15, interquartile range 4–9). The majority of the included studies had a “high” or “unclear” risk of bias, and none of the studies overall had a “low” risk. In conclusion, H2S significantly improves structural and functional outcomes in in vivo animal models of ischemic stroke. However, the level of evidence from preclinical studies is not sufficient to proceed to clinical trials due to the low external validity, high risk of bias, and variable design of existing animal studies.