Although all our subjects were experienced psychophysical observers who were practiced in our task and our stimuli were well separated from the central fixation point, eye movements were a potential confound in this study. We tested the stability of eye fixation for seven of our subjects in separate sessions. Subjects performed an identical task to that used in the EEG experiments while their eye position was monitored with an Eyelink 1000 (SR Research, Chicago, IL) eye tracker sampling at 250 Hz. Viewing was binocular, but only the dominant eye of each subject was tracked. Head position was maintained with a chin rest. Calibration under these conditions is very stable, with high precision and very little drift over the course of an experiment (30–40 arc minutes in our case). Calibration was performed in two stages. First, observers viewed a static cross that was made of 0.1dots placed centrally, T5vertically, and T5and T10horizontally. Observers were instructed to view each dot in turn as the eye tracker was manually adjusted to return a linear readout in response to eye position. Next, observers fixated a dynamic 0.25dot that blinked on for 1500 ms and swept out a 5 5 calibration grid that covered the stimulus space in steps of 5horizontally and 3.5vertically. This grid was used to apply a perspective transformation to the raw eye position data, correcting any non-linearity that may occur near the edges of the display. The nominal standard deviation of this system’s measurements is 0.25-, and we have confirmed this in separate experiments. At the end of the eye tracking experiment, we repeated the second part of the calibration to verify that it was within the nominal standard deviation (Supplementary Figures 1 and 2). Given the size of the fixation point and the precision of the eye tracking system, we found that an ideal observer would maintain fixation within 0.583.9% of the time and 199.6% of the time if they fixated exactly at the edge of the fixation point. All subjects maintained fixation within 0.5for an average of 86.5% of the time and within 1for an average of 99.2% of the time (Supplementary Figure 3, Supplementary Table 1). Both measures indicate that the subjects were able to perform the covert attention task while maintaining fixation on the fix point. Further, the average frequency of saccades ending in the attended grating was 0.48/min during the attention trials, with no more than one saccade for any single 15-s trial.