Aim: To compare the cardiovascular autonomic reflex tests (CARTs) with cardiac sympathetic innervation imaging with I-metaiodobenzylguanidine (MIBG) in patients with type 1 diabetes mellitus (T1DM).Patients and Methods: Forty-nine patients (29 males, mean age 36 ± 10 years, mean T1DM duration 19 ± 6 years) without cardiovascular risk factors were prospectively enrolled. Participants were evaluated for autonomic dysfunction by assessing the mean circular resultant (MCR), Valsalva maneuver (Vals), postural index (PI), and orthostatic hypotension (OH). Within one month from the performance of these tests, patients underwent cardiac MIBG imaging and the ratio of the heart to upper mediastinum count density (H/M) at 4 hours postinjection was calculated (abnormal values, H/M < 1.80).Results: Twenty-nine patients (59%) had abnormal CARTs, and 37 (76%) patients had an H/M_4 < 1.80 ( = 0.456). MCR, PI, Vals, and OH were abnormal in 29 (59%), 8 (16%), 5 (10%), and 11 (22%) patients, respectively. When using H/M_4 < 1.80 as the reference standard, a cutoff point of ≥2 abnormal CARTs had a sensitivity of 100% but a specificity of only 33% for determining CAN.Conclusions: CARTs are not closely associated with I-MIBG measurements, which can detect autonomic dysfunction more efficiently than the former. In comparison to semiquantitative cardiac MIBG assessment, the recommended threshold of ≥2 abnormal CARTs to define cardiovascular autonomic dysfunction is highly sensitive but of limited specificity and is independently determined by the duration of T1DM.