This study examines the effects of sex, socioeconomic status, birth order and birth spacing on the cognitive performance of 12-week-old infants. A brief review of research on neonatal cognitive ability is followed by a description of the study itself. The subjects, 189 three-month-old Caucasian infants (61 first borns, 58 second borns, and 49 third borns) were given a battery of perceptual-cognitive tasks. These tasks included the Corman-Escalona Scales of Object Permanence, the Mental Development Index (MDI) of the Bayley Scales, and an attention task measuring habituation and dishabituation. An analysis of data failed to show any sex or social class differences, although the MDI was affected by birth order, with the first borns showing superior performance. In general, the three measures of perceptual-cognitive ability were unrelated. These data are used to question the notion of a developmentally constant, unchanging, unitary construct of intelligence ('g' factor) in infancy, and therefore the usefulness of a test of infant intelligence. (Author/CM)