Eye movements were measured using binocular infrared VOG (Sensorimotoric Instruments, Berlin, Germany) that samples at 60 Hz with a resolution of 0.1° and a calibrated accuracy of 0.5° in children. The VOG recordings and target position data for saccades, smooth pursuits, horizontal OKN, and VOR were exported and analyzed offline using several analysis programs. A custom interactive program (written by co-author JPK) was used to analyze saccades, smooth pursuit, and OKN data. Each VOG recording was manually edited to remove recording artifacts. Individual saccades in the direction of the target step were analyzed using a settable velocity criterion (>30°/s). Each saccade was then analyzed to extract saccade amplitude, peak velocity, and latency to onset from the target step. Saccade peak velocities were derived from filtered position traces (frequency roll-off −3.0 dB at 15 Hz). Saccade gain was calculated as saccade amplitude/target step amplitude. Analysis of sinusoidal smooth pursuit gain was performed by desaccading the recorded eye position traces based on a settable velocity criterion, and then fitting slow-phase eye velocity to a sinusoid of the same frequency as that of the target motion using a least squares algorithm. Smooth pursuit gain, calculated as the ratio of the fit to desaccaded eye velocity divided by the fit to target velocity, was averaged across cycles. Sinusoidal VOR was analyzed using a similar approach. Eye velocity traces were desaccaded and accumulated for multiple cycles of chair rotation and then fit with a least squares approximation to a sine wave at the frequency of the rotation. The gain, phase, and symmetry of the response were calculated. Optokinetic nystagmus velocity was calculated for each slow phase using a settable velocity criterion to define fast phase start and end, thereby defining the intervening slow-phase eye movement. Optokinetic nystagmus gain was defined as the average velocity of all slow phases divided by the corresponding stimulus velocity.