The original "Bayley Scales of Infant Development" (BSID) (Bayley 1969) have been among the most popular measures of performance and aptitude of infants. In the present study the reliability and construct validity of mental and motor scores on the revised Bayley Scales of Infant Development-II (BSID-II) were investigated. A national sample of 100 children aged 2 to 42 months was used to provide data that were analyzed with second-order factor analysis. The distribution of first-order trace suggests that probably too few first-order factors have been extracted in previous research with the Bayley measures. The findings suggest that the BSID-II assesses multiple aspects of infant behavior. At the second-order level, the mental and motor scores might still emerge as discrete constructs. Results, however, are inconsistent, with discernible mental and motor dimensions in some age groups and not others. It seems that more than two second-order factors must be extracted to fully represent BSID-II score variance. Appendix A presents 16 tables of item difficulty and description data. Appendix B presents 48 tables of reliability analyses. The text contains 34 tables of analysis findings. (Contains 25 references.) (SLD)