Recent reports of transient neurologic deficits have raised concern about the potential toxicity of single-dose spinal 5% lidocaine in 7.5% dextrose. Two cases of volunteers who experienced minor local sensory deficits after slow (60 s) injections of 2 mt 5% lidocaine via Whitacre needles are described. One case was a result of a double injection because of a ''failed'' block. It seemed possible that the neurologic deficit in these cases resulted from neurotoxicity associated with maldistribution of local anesthetic. Using an in vitro spinal model, we investigated drug distribution resulting from injections through side-port spinal needles to determine whether the use of these needles could result in high local concentrations of hyperbaric solutions. A spinal canal model was fabricated using human magnetic resonance measurements. The model was placed in a surgical supine position and filled with lactated Ringer's solution to simulate the specific gravity of cerebral spinal fluid at 22 degrees C. A hyperbaric solution of phthalocyanine blue dye and dextrose (SG 1.042), simulating the anesthetic, was injected through three different needles (27-gauge 4 11/16-in. Whitacre, 25-gauge 3 1/2-in. Whitacre, 25-gauge 3 1/2-in. Quincke). Triplicate injections were done at rapid (2 mL/10 s) and slow (2 mL/60 s) rates, with needle side ports oriented in a sacral and cephalad direction. At slow rates of injection, using 27-or 25-gauge sacrally directed Whitacre needles, injections showed evidence of maldistribution with extrapolated peak sacral lidocaine concentrations reaching 2.0%. In contrast, distribution after slow injection through sacrally directed Quincke needles was uniform. At fast injection rates distribution was uniform for both needle types. These findings are similar to results from small-bore catheter injections in similar models with single-dose peak lidocaine concentrations reaching 2.39%. Our data suggest that sacral needle direction and slow rate of injection with Whitacre needles may predispose to potentially neurotoxic concentrations of intrathecal hyperbaric lidocaine.