A review of literature giving evidence of response preferences, mainly in human Ss, is undertaken. The preferences are taken from experimental work in early psychophysical data, subjective estimates of "chance" sequences, response mechanisms at the threshold, "gambling" situations, probability learning, and the influence of instructions on performance in probability learning. Explanations of response preferences are reviewed and the relationships between subjective uncertainty and preferences are explored with an attempt to give cohesion to a diverse body of experimental evidence.