Weight/aperture (W/A) and weight/square aperture (W/A2) rather than weight (W) previously emerged as possible heaviness models for when a cube is haptically held by thumb-index finger grasp. This is based on the convincing evidence that discernment of a cube's heaviness depends on the integration of sensory information about W and size of finger aperture (A). The present study, therefore, determined which model would be the best predictor for subject-perceived heaviness. Fifteen subjects were asked to judge whether the second of a pair of cubes was heavier, lighter, or identical to the first in 3 x 3 (density x weight) conditions. The subject responses were compared with the expected responses for each of the possible heaviness models as well as W. The percentage of consistent trials, i.e., subject response matched the expected response, was then compared among the three models and conditions within each model. The results indicated that subject responses were significantly more consistent with the expected responses determined from W/A compared with those determined from W or W/A2, suggesting W/A as a best predictor of perceived heaviness by finger-grasp perception. W/A2, however, reflected subject responses almost as accurately as W/A could and was, therefore, discussed as one of possible heaviness models when subjects perceive heaviness with different sensory modality.