The relationship between conceptual complexity and stimulus saliency was explored in a 3 x 4 factorial design using 144 undergraduates. Levels of complexity were represented by varying the ratio of relevant-to-irrelevant dimensions (1:3, 2:2, and 3:1). The saliency factor varied according to the discriminability between each attribute within relevant and irrelevant dimensions. Concept stimuli consisted of four building elements, each containing five attributes that varied in size. The curvilinear relationship found among the complexity conditions for trials and errors cancelled out the differences in difficulty between the four levels of saliency except where there was an equal proportion of relevant and irrelevant dimensions. (Author)