Compared the performance of 33 American and 30 Arabic 6- and 7-yr-olds on 2 individually administered 3-item experimental test forms requiring pictorial depth perception of scenes from American and Arabic cultures. Higher mean scores were obtained for (a) American Ss on the American version, (b) Arabic Ss on the Arabic form, (c) American Ss on the American than on the parallel Arabic version, and (d) Arabic Ss on the Arabic than on the American form. For American Ss, internal-consistency reliability estimates were .90 and .86 for scores on the American and Arabic versions, respectively; for Arabic Ss, the corresponding estimates were .85 and .80. Statistical results support the conclusion that differences in the cultural backgrounds of Ss relative to their familiarity with pictorial content of test items incorporating size and distance cues contributed to significant differences in average level of test performance on the 2 forms. Implications for valid interpretation of test scores derived from pictorial items portraying objects specific to a given culture are discussed.