OBJECTIVE To assess the prevalence and severity of autonomic dysfunction (AD) in patients with systemic lupus erythematosus (SLE). METHODS Fifty-nine consecutive patients with SLE and 97 healthy controls were assessed for AD using 4 noninvasive tests: heart rate responses to the Valsalva maneuver (Val), maximum-minimum heart rate (MM) response to deep breathing, heart response to standing up (30:15 ratio), and changes in blood pressure to sustained handgrip (Hand). AD was categorized as incipient, definite, severe, or atypical according to criteria proposed by Ewing. Disease characteristics, comorbid conditions, disease activity index [Mexican Systemic Lupus Erythematosus Disease Activity Index (Mex-SLEDAI)], cardiac complaints, and drug history were also documented. Laboratory analysis included rheumatoid factor, antinuclear antibodies, anti-dsDNA antibodies, and anticardiolipin antibodies (aCL). RESULTS Forty-eight percent of the patients with SLE had one or more abnormal tests compared with 33% of the controls (p=0.05); differences were greater when the severity of the neuropathy was considered: 12 (21%) patients with SLE had definite or severe AD using Ewing's criteria, compared with only 2 (2%) controls (p < 0.001). No statistical associations were observed between AD and age, disease duration, hypertension, Mex-SLEDAI, anti-dsDNA, aCL antibodies, or other drugs. CONCLUSIONS The prevalence of AD in SLE using noninvasive tests is high. One-fifth of our patients had definite or severe AD. No significant associations were observed between AD and clinical or serological variables of disease activity. Prospective studies are needed to determine the potential effect of AD in the morbidity of patients with SLE.