Uncertainty in various forms plagues our interactions with the environment. In a Bayesian statistical framework, optimal inference and prediction, based on unreliable observations in changing contexts, require the representation and manipulation of different forms of uncertainty. We propose that the neuromodulators acetylcholine and norepinephrine play a major role in the brain's implementation of these uncertainty computations. Acetylcholine signals expected uncertainty, coming from known unreliability of predictive cues within a context. Norepinephrine signals unexpected uncertainty, as when unsignaled context switches produce strongly unexpected observations. These uncertainty signals interact to enable optimal inference and learning in noisy and changeable environments. This formulation is consistent with a wealth of physiological, pharmacological, and behavioral data implicating acetylcholine and norepinephrine in specific aspects of a range of cognitive processes. Moreover, the model suggests a class of attentional cueing tasks that involve both neuromodulators and shows how their interactions may be part-antagonistic, part-synergistic. The brain must handle uncertainty in various forms when making inferences about the world and predicting the future. Bayesian theory provides a quantitative framework for this, and has been applied to cognitive phenomena in perception, attention, and sensorimotor learning. The Bayesian framework formalizes the notion that optimal inference and learning depend critically on representing and processing the various sorts of uncertainty associated with a behavioral context. A context consists of stable statistical regularities that relate environmental entities to each other and to our sensory and motor systems. These relationships allow inferences to be made about imperfectly observed aspects of the environment based on prior observations, which serve as predictive cues. According to Bayesian theory, uncertainty about the behavioral context should suppress the use of assumed cues for making inferences compared with direct sensory information, but boost learning about the lesser known predictive relationships within the current behavioral context. Expected uncertainty arises from known unreliability of predictive relationships within a familiar environment, while unexpected uncertainty is induced by gross changes in the environment that produce sensory observations strongly violating top-down expectations. For instance, the decision to bring an umbrella in the morning involves considering various potentially conflicting sources of information. For someone who regularly views the weather forecast, the typical chance of a misforecast constitutes a form of "expected uncertainty," while a substantial drop in forecast reliability would induce "unexpected uncertainty." The neural realization of expected and unexpected uncertainty signals should suppress top-down, expectation-driven information relative to bottom-up, sensory-induced signals, as well as promote learning about the context. Experimental evidence suggests that the cholinergic and noradrenergic systems satisfy these conditions, with acetylcholine involved in expected uncertainty and norepinephrine in unexpected uncertainty. Across primary sensory cortices, ACh and NE selectively suppress intracortical and feedback synaptic transmission, while sparing or even boosting thalamocortical processing. This suggests that higher ACh and NE levels lead to a suppression of top-downUncertainty in various forms plagues our interactions with the environment. In a Bayesian statistical framework, optimal inference and prediction, based on unreliable observations in changing contexts, require the representation and manipulation of different forms of uncertainty. We propose that the neuromodulators acetylcholine and norepinephrine play a major role in the brain's implementation of these uncertainty computations. Acetylcholine signals expected uncertainty, coming from known unreliability of predictive cues within a context. Norepinephrine signals unexpected uncertainty, as when unsignaled context switches produce strongly unexpected observations. These uncertainty signals interact to enable optimal inference and learning in noisy and changeable environments. This formulation is consistent with a wealth of physiological, pharmacological, and behavioral data implicating acetylcholine and norepinephrine in specific aspects of a range of cognitive processes. Moreover, the model suggests a class of attentional cueing tasks that involve both neuromodulators and shows how their interactions may be part-antagonistic, part-synergistic. The brain must handle uncertainty in various forms when making inferences about the world and predicting the future. Bayesian theory provides a quantitative framework for this, and has been applied to cognitive phenomena in perception, attention, and sensorimotor learning. The Bayesian framework formalizes the notion that optimal inference and learning depend critically on representing and processing the various sorts of uncertainty associated with a behavioral context. A context consists of stable statistical regularities that relate environmental entities to each other and to our sensory and motor systems. These relationships allow inferences to be made about imperfectly observed aspects of the environment based on prior observations, which serve as predictive cues. According to Bayesian theory, uncertainty about the behavioral context should suppress the use of assumed cues for making inferences compared with direct sensory information, but boost learning about the lesser known predictive relationships within the current behavioral context. Expected uncertainty arises from known unreliability of predictive relationships within a familiar environment, while unexpected uncertainty is induced by gross changes in the environment that produce sensory observations strongly violating top-down expectations. For instance, the decision to bring an umbrella in the morning involves considering various potentially conflicting sources of information. For someone who regularly views the weather forecast, the typical chance of a misforecast constitutes a form of "expected uncertainty," while a substantial drop in forecast reliability would induce "unexpected uncertainty." The neural realization of expected and unexpected uncertainty signals should suppress top-down, expectation-driven information relative to bottom-up, sensory-induced signals, as well as promote learning about the context. Experimental evidence suggests that the cholinergic and noradrenergic systems satisfy these conditions, with acetylcholine involved in expected uncertainty and norepinephrine in unexpected uncertainty. Across primary sensory cortices, ACh and NE selectively suppress intracortical and feedback synaptic transmission, while sparing or even boosting thalamocortical processing. This suggests that higher ACh and NE levels lead to a suppression of top-down