Finger movements in the sagittal plane mainly consist of flexion and extension about the metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints. The purpose of this study was to develop a biomedical finger model and to validate it by comparing finger strength and muscular forces in static exertions, which predicated from the model and measured in experiments. Two kinematic finger models were developed: one was with the assumption of constant tendon moment arms, and the other was with the assumption of non-constant tendon moment arms. Equations of static equilibrium were derived for these finger models using the principle of virtual work. Equations of static equilibrium for the finger models were indeterminate since only three equations were available for five unknown variables (forces). By reducing the number of variables based on information in the literature on muscular activities in finger movements, the amounts of force which muscles exerted to maintain static equilibrium against an external load were computed from the equilibrium equations. The muscular forces were expressed mathematically as functions of finger positions, tendon moment arms, lengths of phalanges, and the magnitude and direction of external load. Equations of muscular forces were used to predict external finger strength and to compute internal muscular forces in static exertions against an external load. Computer simulations were performed to compute finger strengths and muscular forces at various finger positions and directions of force exertions. For this, finger positions were controlled, and lengths of phalanges were measured. Experiments were performed to measure finger strengths and muscular activity levels in submaximal contractions. Muscular activity levels were estimated by ratios of standardized EMG amplitude to exerted force. Measurements were taken in combinations of four finger positions and four directions of force exertions. Validation of the biomechanical finger models was done by comparing the results of computer simulations and experiments. Significant differences were found between the predicted and measured finger strengths. However, the trends of finger strengths with respect to finger positions were similar in both the predicted and measured. Trends of variations in predicted and measured muscular activity levels were not different from each other. These findings indicate that the finger models and the procedure to predict finger strengths and muscular forces were correctly developed. / Ph. D.