OBJECTIVE To study the clinical usage of sweep pattern visual evoked potential (SPVEP) acuity in children's visual development periods and compare the amplitude-spatial frequency (A-SP) function regression method with the amplitude-logarithm of the visual angle (A-logVA) function regression method in evaluating the SPVEP acuity of children, especially those who have poor visual acuities. METHODS Twenty-six eyes of 26 amblyopic children (ages ranged from 3 to 12 years; mean age±standard deviation 6.69±1.74 years) and 31 eyes of normal children whose ages were matched with the amblyopic group were involved in this study. SPVEP acuity was recorded with GT-2000 NV (Guote Medical Apparatus Ltd., China) using sinusoidally modulated horizontal gratings with 10 different spatial frequencies from 0.99 to 12.89 cycles per degree to stimulate the retina. The averaging responses were displayed with the discrete Fourier transformation method. SPVEP acuity was assessed by both the A-SP function regression method and the A-logVA function regression method. The logarithm of minimal angle of resolution (logMAR) chart was used to obtain logMAR visual acuity. RESULTS In the normal group, logMAR acuity calculated by both the A-SP and A-logVA function regression methods had a significant correlation with SPVEP acuity. The average value of SPVEP acuity (by A-logVA) was closer to logMAR acuity. The difference of mean values between logMAR acuity and SPVEP acuity was significant in both regression methods. In the amblyopic group, it was SPVEP acuity (by A-logVA) that had a significant correlation with logMAR acuity, whereas the result was not significant when calculated by the A-SP function regression method (p=0.515). The average value of SPVEP acuity (A-SP) was closer to logMAR acuity. The difference of mean values between logMAR acuity and SPVEP acuity (A-logVA) was significant; however, when compared with SPVEP acuity (A-SP), it was not significant (p=0.174). In addition, SPVEP acuity may be overestimated or underestimated when it is compared with different logMAR visual acuities. CONCLUSIONS SPVEP could be used to evaluate the visual acuity for normal children or those with poor visual acuity. Moreover, the A-logVA function regression method was more accurate than the A-SP function regression method in evaluating SPVEP acuity.