<p>Cognitive control deficits are associated with problematic use of short-form video (PUSV), generally based on self-reported and behavioral data. The drift–diffusion model (DDM) provides a computational framework to analyze evidence accumulation processes underlying cognitive control deficits. This study employs an eye-tracking task to examine the association of cognitive control deficits with PUSV and the relative contribution of its underlying evidence accumulation processes. Data were collected from 104 participants using the Interleaved Pro- and Anti-Saccade Task (IPAST) and the problematic short-form video usage test (PSVUT). We applied the DDM to integrate IPAST reaction times and accuracy rates for deriving computational parameters—drift rate, decision threshold, and non-decision time—to quantify the latent cognitive control processes. In IPAST, the anti-saccade condition exhibited lower accuracy rates and longer reaction times than the pro-saccade condition. Drift rates were significantly lower for anti-saccade than pro-saccade conditions per DDM analysis. Correlation analysis showed that PSVUT scores were negatively associated with accuracy rates, drift rates, and decision threshold in the anti-saccade condition. The results suggest that people with PUSV tend to have greater cognitive control deficits, because of slower evidence accumulation and greater response impulsiveness according to the DDM. The findings drawn from the computational approach and eye-tracking data provide strong behavioral evidence for the risk effect of cognitive control deficits on PUSV and thus practical implications for interventions.</p>