Vernier acuity and grating acuity were measured longitudinally starting at 1 or 2 months of age in 22 infants, using a two-alternative, forced-choice preferential looking technique. For vernier acuity, the motion-sound display was employed. For grating acuity, a preferential looking method was employed. Steps of the stimulus (vernier offset and spatial frequency of the grating) and procedures were basically identical between the two acuity tests. The range of stimuli was set so as to compare the two acuities at younger ages. Results show: vernier acuity is less than grating acuity at 11-12 weeks of age or younger, and the developmental rate of vernier acuity is greater than that of grating acuity in the first half-year of life. To interpret the data, it was speculated that: the mean sampling distance (center-to-center distance between receptive fields) may influence vernier acuity more than grating acuity, whereas the size of the receptive field may influence grating acuity more than vernier acuity: when the mean sampling distance is large relative to the size of the receptive field, vernier acuity may be less than grating acuity. Thus, the neonatal visual system, just as the visual system in the periphery and in strabismic amblyopia, may be characterized by spatial undersampling.