This paper reports two studies which compared scores obtained on the Slosson Intelligence Test (SIT) and Peabody Picture Vocabulary Test (PPVT) and investigated the hypothesis that the representational level of the stimulus items in the PPVT is inappropriate for preschool children regardless of socioeconomic background. In the first study, the PPVT and SIT were administered individually to a total of 18 Head Start (HS) children of median age 5 years 2 months. In the second study 18 HS and 18 middle income (MI) 4- and 5-year-old children were asked to label stimuli shown in three representational forms: objects, colored pictures and black and white line drawings. Both MI and HS children labeled the objects more successfully than the pictures, but this discrepancy was greater for the HS children. It was confirmed that the HS children obtained lower scores on the PPVT than on the SIT, and it is suggested that this may be due to the highly representational stimuli of the PPVT. A training effect was found in both the HS and MI groups; performance was enhanced by the presentation of objects before pictures. The possibility of cultural bias in preschool screening tests is discussed and it is suggested that such bias may be reduced by: (1) using concrete objects in preschool stimulus labeling tasks, (2) ordering stimulus presentation from concrete to highly representational forms, and (3) considering the appropriateness of the SIT compared with the PPVT. (Author/GO)