Ss responded to a 1,000-Hz tone of 50, BO, or 100 dB. Catch trial conditions were none! blan~~ials, a red light, a noise, and an 1,BOO-Hz tone. Auditory catch. signals were o.f the same mteJ;tsltles. RT distributions in the first three conditions were well descnbed by a family of exponential growth functions dependent upon stimulus intensity and by the parameters of normal criterion distribu~ions dependent upon catch trial conditions and between-session variability. Perform~ce in the auditory catch trial conditions was not dependent upon the same set of sensory growth functions. Performa!Ice m these conditions was described by a two-dimensional analysis of information transmitted as a function of time and interpreted in terms of variable criterion theory: The speed-accuracy~rad~off in t~is situation appears to depend upon differential rates of growth of intensity and associative information and the criterion used in responding to this information.In the context of a decision theory approach to reaction time (RT), catch trial effects may be viewed as resulting from variation in the decision criterion. Grice (1972b) has supported this interpretation in a quantitative analysis of an experiment by LaBerge (1971). Grice's theory of response evocation assumes that, following stimulus onset, sensory information (V) grows in strength according to some continuous function, the rate of growth depending upon stimulus intensity. In conditioning, and presumably in choice situations, there is also an associative information component which grows in a somewhat slower fashion (Grice, 1972a). When this sensory (and, when applicable, associative) growth reaches the level of the S's decision criterion or reaction threshold (T), the response occurs. Thus, in RT the latency is determined by the rate of growth of V and by the level of the criterion. The criterion is assumed to depend upon the requirements of the task; set, attentional, and motivational factors; and individual differences. Under homogeneous conditions of performance, it is assumed to be a normally distributed random variable with mean, T, and standard deviation, a. Both of these parameters are influenced by experimental conditions and individual differences. The value determining the probability of response at a particular latency is E =