GM1 ganglioside decreases the severity of ischemic brain lesions in experimental models, although the mechanism is uncertain. In clinical trials involving patients with stroke, efficacy has been reported in some but not in others. However, some of the latter also showed efficacy after analyses not planned before the trial began. Analyses of the trials done to date revealed design differences sufficiently large so as to preclude meta-analysis of the results. Moreover, flaws in these studies may account for some of their failure to demonstrate that GM1 therapy is efficacious in ischemic stroke. Several of these flaws are discussed, including small sample size; attrition of the study cohort; inclusion of stroke severity and type that made demonstrations of a beneficial effect difficult; use of inappropriate clinical and outcome measuring instruments; delay in enrollment; inappropriate statistical analyses; inadequate dose; inappropriate route of administration; a too short duration of treatment. Improvements in these design features in future clinical trials of GM1 may yet demonstrate efficacy of this drug in acute ischemic stroke.