We assessed the validity of clinical diagnosis of topography (CDT) and etiology (CDE) in 100 consecutive patients who were examined within 12 h of stroke onset by a stroke team (1 neurologist with interest in stroke, 1 fellow, 1 resident). The team established a first- and a second-choice CDT (among 28 possibilities) and CDE (among 11 possibilities). CDT was compared with CT and MRI findings (2 to 13 days after stroke) in all patients. Potential causes of stroke were assessed with the standard protocol of investigations of the Lausanne Stroke Registry. Stroke diagnosis was confirmed in all 100 patients and all 4 hemorrhages were diagnosed before CT/MRI. First-choice CDT (in 68) or second-choice CDT (in 20) was in agreement with CT/MRI in 88 patients. CDT was wrong in 4 patients and CT/MRI showed no lesion in 8 patients. First-choice CDE (in 70) or second-choice CDE (in 4) remained the most likely etiology in 74 patients, while the cause of stroke remained unknown in 24 patients. Thus, first- or second-choice CDT and CDE were confirmed in over 95% of the cases in which final diagnoses could be made. Our findings show that early CDT and CDE may be extremely accurate, and suggest that in acute stroke trials, investigations may not be compulsory for subgroup allocation before randomization.