A method for direct catheterization of the lumbar subarachnoid space has recently been developed by Storkson et al. (1996) (J Neurosci Methods 1996;65:167-172) that may potentially improve upon the widely used method of Yaksh and Rudy (1976) (Physiol Behav 1976;17:1031-1036). This 'catheter-through-a-needle' technique inserts the catheter between lumbar vertebrae 5 (L5) and 6 (L6), which has been shown to reduce neurological impairment and post-surgical deaths. However, employing this technique allows the external portion of the chronic indwelling catheters to be easily damaged, resulting in approximately 50% attrition within 4 days after surgery. Therefore, we developed an easy and inexpensive method for protecting the external portion of the catheter that enhances catheter viability beyond 14 days after catheter implantation while also maintaining a low injection volume (8 microl). Moreover, this modification does not significantly alter the implantation methods developed by Storkson et al. (1996) (J Neurosci Methods 1996;65:167-172) and allows for more optimal catheter materials to be incorporated. Chronically implanted catheters (n = 70) with the external portion of the catheter protected, resulted in 4% attrition 7 days after surgery and 11% attrition 14 days after surgery. Approximately 5.5% of animals implanted showed very mild and transient neurological impairment.