The purpose of this paper is to demonstrate the superiority of the Ame
rican Spinal Injury Association motor level (ML) and upper extremity m
otor score (UEMS) to the neurological level (NL) in determining self c
are function in motor complete tetraplegia. Fifty subjects with trauma
tic motor complete tetraplegia, NL C4-C8, were evaluated at admission
and 12 months post injury. At both time periods NL, ML, and UEMS were
determined. At 12 months, reported ability to perform six feeding acti
vities of the Quadriplegia Index of Function (QIF) were documented. Sp
earman correlations of the NL, BML, WML, UEMS, and feeding QIF scores
were conducted, and results were compared with t tests for significant
differences. Both the best and worst ML were more highly correlated t
o the UEMS than was the NL (0.96 and 0.96 vs 0.66, P < 0.001). The bes
t and worst ML were more highly correlated to the QIF feeding score th
an was the NL (0.74 and 0.72 vs 0.56, P < 0.05). The UEMS had the high
est correlation to the QIF feeding score, 0.78. These results suggest
that the NL is an imprecise descriptor of the impairment in SCI, and i
s therefore a poor predictor of the resultant disability. The ML and t
he UEMS better reflect the severity of impairment and disability after
motor complete tetraplegia.