Manual dexterity is a skill frequently evaluated in rehabilitation to estimate hand function. Several tests have been developed for this purpose, including the Box and Block Test (BBT) that measures gross manual dexterity. The goal of the present study was to verify the test-retest reliability and construct validity of the BBT with subjects aged 60 and over with upper limb impairment. The second objective of this research was to develop normative data based on a random sample of healthy elderly community-living individuals. The results show that the test-retest reliability is high (intraclass correlations coefficients of 0.89 to 0.97) and the validity of the test is shown by significant correlations between the BBT, an upper limb performance measurement and a functional independence measurement. The norms will help rehabilitation clinicians to differentiate better between real difficulties and those that may be attributed to normal aging.