Observational study data have suggested that neurologists preferentially selected younger patients with generally few traditional stroke risk factors for homocysteine testing. The aim of this study was to corroborate observational data by assessing neurologists' reported practices for the detection and treatment of hyperhomocysteinemia in patients with acute ischemic stroke. All academic neurology faculty (n = 49), trainees (neurology residents/fellows, n = 53), and a random sample of community-based neurologists (n = 66) throughout North Carolina were surveyed regarding their homocysteine testing and treatment practices. Nine neurologists were ineligible because they had either retired or moved out of state. A total of 93 of the remaining 159 surveys (58.5%) were completed (response rates: faculty, 74%; trainees, 47%; community-based neurologists, 56%). Patients age < 50 years (n = 52; 63%) and the absence of traditional stroke risk factors (n = 57; 70%) were the most common factors cited as prompting homocysteine testing. The homocysteine level threshold for treatment varied independently by practice type (faculty: median, 14 mumol/L; range, 6-16 mumol/L; trainees: median, 14.5 mumol/L; range, 10-20 mumol/L; community-based: median, 10.4 mumol/L; range, 7-15 mumol/L; P = .01), the number of stroke patients evaluated during a typical week (Spearman's r = .32; P = .034), and year of training completion (Spearman's r = .41; P = .003). About half (51%) treat elevated homocysteine with a combination of folate, vitamin B(12), and vitamin B(6). Consistent with observational data, the majority of the neurologists surveyed report that they select young stroke patients who lack traditional stroke risk factors for homocysteine testing. Thresholds for treatment varied between community-based and academic neurologists and correlated with the physicians' stroke patient volume.