To make the point that infant emotions are more dynamic than suggested by Differential Emotions Theory, which maintains that infants show the same prototypical facial expressions for emotions as adults do, this paper explores two questions: (1) when infants experience an emotion, do they always show the corresponding prototypical facial expression?; and (2) do the facial configurations of the AFFEX coding system (described by Izard, Dougherty, and Hembree) for anger, sadness, and pain really express discrete emotions? To support the argument, the results of laboratory studies and naturalistic observations are cited. Several studies of babies' facial expressions have shown that babies do not always produce expressions of fear or surprise in situations that normally induce those emotions. The lack of situational specificity (the expression of an emotion in a situation in which the emotion is assumed to be present, and the lack of its expression in other situations) for anger, sadness, and pain found by some studies suggests that the expressions may reflect a general form of negative affect. Studies that involved observer judgments of infant facial expressions found that judgments were less accurate than expected. The report concludes that the AFFEX configurations are probably not expressions of discrete anger, sadness, and pain, and offers three possible explanations of the relationship between infants' emotion and expression: (1) infants have discrete emotions of pain, anger, and sadness and emotion-specific facial expressions, although researchers have not yet discovered what those expressions are; (2) infants show the same expressions for anger, sadness, and pain; and (3) infants do not have discrete negative emotions, but do have a less differentiated state of distress. (TJQ)