BACKGROUND: Recent evidence suggests that transient neurologic symptoms commonly follow lidocaine spinal anesthesia. However, information concerning factors that affect their occurrence is limited. Accordingly, to evaluate many potential risk factors, the authors undertook a prospective, multicenter, epidemiologic study. METHODS: On a voluntary basis, anesthetists at 15 participating centers forwarded a data sheet on patients who had spinal anesthesia to a research nurse blinded to the details of anesthesia and surgery. A subset was randomly selected for follow-up. The pressure [corrected] of transient neurologic symptoms, defined as leg or buttock pain, was the principal outcome variable. Logistic regression was used to control for potential confounders, and adjusted odds ratios and confidence intervals were used to estimate relative risk. RESULTS: During a 14-month period, 1,863 patients were studied, of whom 47% received lidocaine, 40% bupivacaine, and 13% tetracaine. Patients given lidocaine were at higher risk for symptoms compared with those receiving bupivacaine (relative risk, 5.1; 95% CI, 2.5 to 10.2) or tetracaine (relative risk, 3.2; 95% CI, 1.04 to 9.84). For patients who received lidocaine, the relative risk of transient neurologic symptoms was 2.6 (95% CI, 1.5 to 4.5) with the lithotomy position compared with other positions, 3.6 (95% CI, 1.9 to 6.8), for outpatients compared with inpatients, and 1.6 (95% CI, 1 to 2.5) for obese (body mass index >30) compared with nonobese patients. CONCLUSIONS: These results indicate that transient neurologic symptoms commonly follow lidocaine spinal anesthesia but are relatively uncommon with bupivacaine or tetracaine. The data identify lithotomy position and outpatient status as important risk factors in patients who receive lidocaine. Among other factors postulated to increase risk, obesity had an effect of borderline statistical significance, whereas age, sex, history of back pain, needle type, and lidocaine dose and concentration failed to affect risk.