The effect of racial characteristics of a model on modeling behaviour was measured by Aboriginal and white Australian boys' responses to three visual recognition tasks performed by Aboriginal and white models. Analysis of variance of data derived from the choice theory of Luce (1963) indicated that Ss changed their response bias to the test stimuli depending on the race of the model. Task differences were also culturally specified by bias scores. Race of model and type of task did not affect Ss' accuracy of responding. The use of choice theory is considered later in relation to experimenter effects in cross-cultural research.