In an effort to examine the effects of institutional racism on Native American preschoolers, a modified version of the Porter doll play and storytelling test was administered to 117 American Indian children (aged 3-5 1/2 years) and a control group of 95 white children comparable in age and geographical distribution. Separated into sub-groups indicating pre-school environmental differences, the Native American sample was derived from: urban white-oriented communities (N=21); reservations with white teachers and Native American aides (N=75); and reservations having no preschool facilities so that the children did not attend preschool (N=21). Employing two sets of dolls differing only in racial representation, a storytelling game was conducted wherein the children selected from a pair of matched dolls, one doll to represent each character in the story. Racial preference was measured by the number of times a child selected a doll of his own racial group for a positive stereotype and by the number of times he selected a doll of the other race for a negative stereotype. Results were: American Indian children scored lower than whites on the self-identification measure and on the test of racial preference; reservation children not attending preschool scored considerably higher than children in the other two groups; the difference between reservation non-attenders and urban dwelling attenders was considerably larger than the difference between whites and reservation non-attenders; and reservation attenders scored better than urban attenders. (JC)