A method was developed for administering intrathecal pharmacotherapy in a rat model of spinal cord injury. The eects of intrathecal administration of nimodipine on spinal cord blood¯ow (SCBF) and evoked potentials (EPs) were measured in the normal and injured spinal cord. It had previously been shown that systemic nimodipine caused severe hypotension after spinal cord injury. After baseline SCBF and EPs, 15 uninjured rats were blindly allocated to one of three groups: one placebo group (n=5); and two groups with intrathecal nimodipine, 0.05 mg/kg (n=5), or 0.2 mg/kg (n=5). Ten other rats received a 35 g acute clip compression injury of the spinal cord for 1 minute and, were allocated to one of two groups: placebo (n=5); and intrathecal nimodipine 0.05 mg/kg (n=5) given 60 min after injury. In the uninjured groups, neither 0.05 nor 0.2 mg/kg of nimodipine increased SCBF during, or 30 min after, intrathecal infusion. However, the mean arterial blood pressure (MABP) decreased signi®cantly to 69.73.1% after the infusion of 0.2 mg/kg nimodipine and did not recover by 98 min. In all three groups of uninjured rats, the amplitude of the cerebellar EP was decreased 30 min after infusion. After spinal cord injury, there were signi®cant decreases in MABP, SCBF and EP amplitude in both placebo and treatment groups, but there was no therapeutic bene®t from nimodipine. Thus, intrathecal infusion of nimodipine did not prevent the hypotension encountered with systemic administration and exerted no bene®cial eect on SCBF or EPs after acute spinal cord injury.