Background and Objectives: Comparisons of swallowing physiology between coronavirus disease 2019 (COVID-19) pneumonia and clinically diagnosed aspiration pneumonia (AP) have largely relied on ordinal scales, leaving etiology-specific biomechanical profiles unclear. We quantitatively compared videofluoroscopic swallowing study (VFSS) measures of pharyngeal residue and clearance mechanics to identify differential dysphagia phenotypes. Materials and Methods: This single-center retrospective cohort study included 50 adult inpatients with pneumonia (COVID-19, n = 25; AP, n = 25) who underwent VFSS for suspected dysphagia. COVID-19 pneumonia was laboratory-confirmed, and AP was clinically diagnosed after negative tests for severe acute respiratory syndrome coronavirus 2. Blinded ImageJ analysis examined the first standardized semisolid yogurt swallow (International Dysphagia Diet Standardisation Initiative level 4). Primary outcomes were Normalized Residue Ratio Scale values for the valleculae (NRRSv) and piriform sinuses (NRRSp); secondary outcomes were upper esophageal sphincter (UES) opening width and epiglottic rotation angle. Penetration–Aspiration Scale (PAS) score, hyoid displacement, and pharyngeal transit time were exploratory. Results: Baseline characteristics were comparable. COVID-19 pneumonia showed higher NRRSv (0.20 [0.12–0.56] vs. 0.13 [0.00–0.20]; p = 0.01). NRRSp was numerically higher but not statistically significant (0.12 [0.00–0.43] vs. 0.00 [0.00–0.17]; p = 0.07). COVID-19 pneumonia also showed smaller UES opening width (5.08 ± 2.48 vs. 6.50 ± 2.01 mm; p = 0.03) and reduced epiglottic rotation angle (66.0 [29.0–80.8] vs. 93.4 [74.2–100.4] degrees; p = 0.04). No statistically significant between-group difference was detected in PAS-defined airway invasion severity on the standardized semisolid task. Conclusions: These findings suggest an efficiency-predominant dysphagia phenotype in COVID-19 pneumonia, characterized by greater vallecular residue and restrictive clearance-related mechanics on a standardized semisolid task. The results indicate that PAS-defined safety metrics alone may underestimate residue-related dysphagia burden in this population.