Recent editorials and reviews express disillusionment and sharp criticism with the contribution of animal experimental studies to stroke prevention and treatment. The basis for these comments appears to be a frustration with the absence of pharmacologic agents to effectively treat stroke patients despite considerable research efforts. In response to these nihilistic views, Zivin and Grotta (Stroke, 21 (1990) 981-983) have written a poignant rebuttle and Goldstein (Stroke, 21 (1990) 373-374) has outlined the considerable progress which has been achieved in experimental cerebral stroke research. It is our goal in the present discussion to highlight the strengths and to examine the potential pitfalls of stroke models, some of which may have contributed to the equivocal results obtained in testing pharmacologic agents. In addition, based on our own experience with a model of middle cerebral artery (MCA) occlusion in cats, we will describe the benefits of large versus small animal models for experimental stroke studies.