We measured the duration and amplitude of saccades in three normal sub
jects, eight patients with myasthenia gravis, and eight patients with
nonmyasthenic ocular palsies. Saccades were examined at the start of a
repetitive saccade task, after 3 minutes of activity, and 1 minute af
ter administration of edrophonium. The duration of saccades was prolon
ged initially in both myasthenic and nonmyasthenic palsies. Activity d
id not produce significant differences among the three groups in eithe
r the slope of the duration-amplitude relationship or the predicted du
rations of saccades of 5 degrees, 10 degrees, or 15 degrees. However,
durations decreased in myasthenia but increased in nonmyasthenic palsi
es after edrophonium. Much of this decrease in myasthenic saccadic dur
ation was due to reduction in deceleration time, indicating resolution
of intrasaccadic fatigue after edrophonium administration. However, t
he relation of deceleration fraction (deceleration time divided by tot
al duration) to total duration remained constant in all subject groups
. Analysis of saccadic duration is a useful means of interpreting resp
onses to edrophonium because it incorporates data from saccades of a w
ide range of amplitudes into a linear relation between duration and am
plitude.