The goal of the current study was to assess general maturational changes in the ERP in the same sample of infants from 4 to 12 months of age. All participants were tested in two experimental manipulations at each age: a test of facial recognition and one of object recognition. Two sets of analyses were undertaken. First, growth curve modeling with mixed models was used to examine trajectories of development and possible differences in trajectories based on recognition memory (novel versus familiar) and/or stimulus-specific memory (face versus object recognition). Our results suggest that the Pb, Nc and Slow Wave components change significantly in terms of amplitude and latency over the first year of life. Pb amplitude showed a significant non-linear increase over time, whereas Pb latency showed a significant linear decrease over time with a plateau beginning at 10 months. Nc amplitude showed a significant linear decrease over time (i.e. a stronger negative value), whereas Nc latency showed a significant linear decrease over time, with a plateau beginning at 8 months. Second, to relate our findings to those reported in the literature, we examined the effects of memory and stimulus and their combination. Differences between recognition memory and stimulus specific memory were found in the responses to familiar and novel faces and objects for all three components, although the pattern differed across the five ages. These results have implications for future studies that involve the recording of the visual ERP, and point to the advantages of growth curve modeling in examining longitudinal data to account for non-linear development.