Background: Spinal cord injury (SCI) impairs sensorimotor function below the lesion and reshapes supralesional circuits, potentially influencing motor control above the injury. Although upper extremity strength and sensation are clinically normal in paraplegia, it is not known whether supralesional reorganization may produce subclinical alterations in the fine motor skills of the upper extremities. The aim of the study was to compare manual dexterity evaluated through the Purdue Pegboard Test between patients with SCI (PwSCI) and healthy subjects (HS). Methods: We recruited 18 PwSCI with complete paraplegia and 18 age- and sex-matched HS. Participants completed the four subtests of the Purdue Pegboard Test: dominant hand (1), non-dominant hand (2), bimanual (3), and assembly (4). For the first three subtests, a mixed 3 × 2 ANOVA (3 subtests × 2 groups) was performed, whereas for the fourth subtest, an independent samples t-test was performed. Spearman’s rho quantified correlations among subtests and with clinical findings. Results: PwSCI showed full upper extremity muscle strength and sensation. The mean scores of the four subtests were significantly lower in PwSCI than in HS. Unilateral and bimanual subtests correlated with each other in both groups; however, the bimanual subtest was not as well predicted by dominant hand alone, as PwSCI depended more on non-dominant ability. In PwSCI, the assembly subtest strongly depended on dominant, non-dominant, and bimanual scores, whereas this dependency was weaker in HS. The subtests were influenced by ageing in PwSCI. Conclusions: Despite upper extremity muscle strength and sensation being clinically normal, PwSCI showed impaired manual dexterity. This may reflect diminished ascending somatosensory input to supraspinal centres and plastic changes in supralesional motor pathways. These preliminary results open new rehabilitation perspectives for PwSCI.