<p>This study investigates the differences in visual attention between expert and novice soccer referees using a multiple object tracking (MOT) task, with a specific focus on how tracking speed modulates these differences. Fifty-two male referees were divided into expert (currently holding international/national level, <i>n</i> = 26) and novice (currently holding level 1 or below, <i>n</i> = 26) groups. Participants completed an MOT task at three tracking speeds (4°/s, 8°/s, and 12°/s), with accuracy recorded and analyzed via a 2 (Group: expert vs. novice) × 3 (Speed: 4°/s, 8°/s, 12°/s) mixed-design repeated-measures analysis of variance (ANOVA), with Speed as the within-subjects factor and Group as the between-subjects factor. Results showed a significant main effect of tracking speed, <i>F</i>(2, 100) = 451.364, <i>p</i> &lt; 0.001, <i>η</i>_<i>P</i>^<i>2</i> = 0.900, with accuracy decreasing as speed increased, and a main effect of expertise, <i>F</i>(1, 50) = 12.406, <i>p</i> &lt; 0.001, <i>η</i>_<i>P</i>^<i>2</i> = 0.199, with experts outperforming novices. Crucially, a significant Group × Speed interaction was observed, <i>F</i>(2, 100) = 10.096, <i>p</i> &lt; 0.001, <i>η</i>_<i>P</i>^<i>2</i> = 0.168: experts demonstrated superior performance at higher speeds (8°/s and 12°/s), but no significant difference was found at the lowest speed (4°/s). Post-hoc analyses with Bonferroni correction revealed large effect sizes for the group differences at 8°/s (<i>Cohen’s d</i> = 0.935) and 12°/s (<i>d</i> = 1.164), whereas the effect at 4°/s was negligible (<i>d</i> = 0.172). These findings suggest that expert soccer referees show superior MOT performance under higher attentional load, indicating a load-dependent expertise advantage.</p>