Several psychophysical studies in audition and VISion have demonstrated that classical methods of determining thresholds do not provide for an evaluation of the S's response criterion independent of threshold measures (Swets, Tanner, & Birdsall, 1961;Swets, 1961;Green & Swets, 1966). This finding has led to the formulation of the theory of signal detection as well as a re-evaluation of classical results in other modalities. Recently the detection model has been applied with great success in, for example, taste (Linker, Moore, & Galanter, 1964), touch (Rollman, 1968), and pain (Clark & Dillon, 1968). Sensory coding in the olfactory system has not been examined within the context of detection theory.. Many threshold studies in olfaction have been performed to determine which of a number of methods of stimulus production, in terms of apparatus design, produces the most reliable reports of threshold (Wenzel, 1948;Jones, 1955;Cheesman & Kirby, 1959;Stone, 1963). Although the use of an olfactometer produces the most reliable results (Stone, 1963), the sniff method such as that described by Engen (1961), also produces highly consistent results. The sniff method was used in the present investigation because stimuli of the same concentration can easily be reproduced from session to session.Another set of olfactory threshold experiments has been conducted to study adaptation effects, the influence of stimulus parameters on threshold, and odor similarity (Cheesman & Mayne, 1953;Moncrieff, 1956;Stuiver, 1958). Engen (1960) used lowering of absolute threshold as an index of practice effects: in addition, he found that verbal instructions to S influenced the reported threshold, but no attempt was made to analyze sensitivity independent of this response-biasing phenomenon.EXPERIMENT 1 Experiment I was performed to determine whether or not practice in detecting odorous butanol stimuli lowers the apparent threshold. Gibson (1953) has reviewed the effects of practice and training on perceptual judgments. Results from several modalities involving a variety of perceptual tasks suggest that perception can be rendered more acute in the sense that S comes to discriminate smaller differences in stimulation as a function of practice. In olfaction, Engen (1960) has demonstrated clear evidence of practice effects with the forced-choice method of limits. He used six odorants which were presented in the same order to five Ss. Threshold for the first three odorants (amyl acetate, phenylethyl alcohol and heptane) decreased markedly as a function of practice, but no such effect was observed for the remaining three stimuli (heptanal, diacetone alcohol and vanillin). These results suggest a transfer of practice effect between odorants, but they do not indicate the extent to which threshold is lowered for any single odorant.The procedure used in Experiment I, the method of double random, yes-no staircases, was first proposed by Comsweet (1962) as a modification of the classical method of limits. The method of limits is subject to several respon...