Detection and resolution of square patches of sinusoidal gratings were measured in central and peripheral vision (30 degrees horizontal temporal visual field) for high-contrast gratings as a function of the number of cycles in the stimulus. We determined performance in a forced-choice paradigm for a fixed number of stimulus cycles by arranging for stimulus diameter to vary inversely with spatial frequency. For both psychophysical tasks and for both target locations, the psychometric function relating performance to log spatial frequency shifted to higher frequencies without changing slope significantly as the number of cycles in the stimulus was increased. Thus the entire effect could be captured by an analysis of spatial acuity, which increased with increasing number of grating cycles over the range 0.5-6 cycles but remained constant over the range 6-14 cycles. In the central field, resolution acuity and detection acuity were equal regardless of the number of cycles in the stimulus. In the peripheral field, detection acuity exceeded resolution acuity and perceptual aliasing occurred for stimuli in the range 1-14 cycles. From this result we conclude that resolution acuity is sampling limited in the periphery, provided that the stimulus contains at least one full cycle of the grating. Essential features of the results could be accounted for by Fourier analysis of the stimulus.