We studied 25 patients within 6 hours of acute spinal cord injury usin
g magnetically evoked cortical motor evoked potentials (MEPs). The sub
jects included 16 quadriplegics with cervical spine injuries and eight
paraplegics. MEPs were recorded from abductor digiti minimi (ADM), bi
ceps, flexor hallucis brevis, and tibialis anterior muscles on each si
de using appropriate directions of coil current. MEPs were not obtaine
d, either at rest or during attempted voluntary contraction, in patien
ts without preceding clinical evidence of voluntary activation. This w
as the case even for muscles that later had motor recovery after an in
itial paralysis. In comparison with normal controls, MEP thresholds at
rest in ADM and biceps were elevated even for muscles innervated abov
e the level of injury (p < 0.001). On repeat testing, 6 weeks postinju
ry, there was a return toward normal, but a significant threshold elev
ation remained for ADM (p < 0.05). We conclude that in acutely injured
spinal cord patients, magnetically evoked MEPs do not provide useful
information regarding the likelihood of motor recovery.