The authors sought to determine whether errors of action committed by
patients with closed head injury (CHI) would conform to predictions de
rived from frontal lobe theories. In Study 1, 30 CHI patients and 18 n
ormal controls performed routine activities, such as wrapping a presen
t, under conditions of graded complexity. CHI patients committed more
errors even on the simplest condition; but, except for a higher propor
tion of omitted actions, their error profile was very similar to that
of controls. Study 2 involved a subset of patients whose performance i
n Study 1 was within normal limits. When these high functioning patien
ts were asked to perform the routine tasks under still more taxing con
ditions, they, too, committed errors in excess of the control group. A
ccounts based on frontal mechanisms have a difficult time explaining t
he overall pattern of findings. An alternative based on limited-capaci
ty resources is suggested.