Saccadic eye movements are essential for redirecting the fovea at different visual targets. In adults and children saccades are remarkably stereotyped. Peak velocity and duration of saccades are a simple function of saccade amplitude called the 'main sequence'. Saccades that are substantially slower than normal often reflect disease of the brain stem saccade generator but may also be associated with diseases of higher level structures including the cerebral hemispheres and superior colliculus. However, little is known about the speed of saccades in infancy. A single previous study reported that infant saccades may be similar to or slower than those of adults, but few saccades were recorded. The present study re-examined this issue with the technique of measuring optokinetic (OKN) quick phases, which are readily elicited from healthy and sick infants, with a view to using saccade speed as a quantitative neurological measure. We measured the duration and peak velocity of saccades (main sequence) using direct-current electro-oculography from OKN quick phases in 18 infants (nine males, nine females) aged 2 to 18 months (mean age 8 mo [SD 4]) and seven adult comparison participants (four males, three females; age range 21-32 y, mean age 27 y [SD 3]). All infant saccades showed typical relationships between duration, peak velocity, and amplitude. Overall, there was no statistically significant difference between adult and infant main sequences for duration or peak velocity. However, the differences in the main sequence for duration almost reached significance (p=0.051) for infant saccades being faster than adults. Individual differences were also present, and some infants produced saccades faster than adults, but not slower. There was no significant age trend. We conclude that measuring saccade speed is practicable in the young infant. From the age of at least 2 months, infants generate saccades with speeds similar to or slightly higher than those of adults.