The article reviews measures of emotion, focusing on whether different types of emotion-evocative stimuli are associated with discrete and invariant patterns of responding in each response system (experiential, physiological, and behavioural). It examines how these responses are structured and whether they converge across systems. The evidence suggests that emotional responding is better described by dimensions (valence, arousal, approach-avoidance) rather than discrete states. Each response system has unique sources of variance, limiting convergence across measures. Therefore, no single measure is a "gold standard," and all measures are relevant to understanding emotion. Self-report measures of emotion are more valid when they relate to current experiences rather than past or future ones. However, they can be biased due to factors like social desirability and alexithymia. Dimensional frameworks, such as valence and arousal, capture more variance in self-reported emotional responses than discrete dimensions. Autonomic measures (e.g., heart rate, skin conductance) are influenced by arousal and valence but not necessarily by discrete emotions. While some studies suggest autonomic specificity, a meta-analysis found inconsistent results. Instead, broader dimensions like arousal and valence are more consistent with autonomic responses. Multiple autonomic measures may better capture discrete emotional states. Startle response magnitude is a measure of emotional valence, with larger responses to negative stimuli and smaller responses to positive ones. However, it does not assess discrete emotions. Startle responses are influenced by arousal and valence, and are more sensitive to high-arousal stimuli. Brain states, as measured by EEG and neuroimaging, show that left-hemisphere activation is associated with approach-related emotions, while right-hemisphere activation is associated with avoidance-related emotions. Neuroimaging studies suggest that specific brain regions (e.g., amygdala, insula, prefrontal cortex) are linked to different emotions, but findings are inconsistent and often confounded by induction methods. Behavioural measures, such as vocal characteristics and facial expressions, are sensitive to emotional valence. Vocal pitch and amplitude are associated with arousal, while facial expressions (e.g., Duchenne smiles) are linked to positive emotions. However, facial behaviours may not always reflect discrete emotions due to factors like culture and audience presence. Whole-body behaviour, such as posture, is associated with certain emotions (e.g., pride, embarrassment). These behaviours may be more relevant to social-status-related emotions. In conclusion, emotional responses are better understood through dimensional frameworks rather than discrete states. Different measures of emotion are sensitive to different dimensions, and no single measure is sufficient to capture the complexity of emotion. The review supports the idea that emotional states are best understood through a dimensional perspective, with multiple measures providing complementary insights.The article reviews measures of emotion, focusing on whether different types of emotion-evocative stimuli are associated with discrete and invariant patterns of responding in each response system (experiential, physiological, and behavioural). It examines how these responses are structured and whether they converge across systems. The evidence suggests that emotional responding is better described by dimensions (valence, arousal, approach-avoidance) rather than discrete states. Each response system has unique sources of variance, limiting convergence across measures. Therefore, no single measure is a "gold standard," and all measures are relevant to understanding emotion. Self-report measures of emotion are more valid when they relate to current experiences rather than past or future ones. However, they can be biased due to factors like social desirability and alexithymia. Dimensional frameworks, such as valence and arousal, capture more variance in self-reported emotional responses than discrete dimensions. Autonomic measures (e.g., heart rate, skin conductance) are influenced by arousal and valence but not necessarily by discrete emotions. While some studies suggest autonomic specificity, a meta-analysis found inconsistent results. Instead, broader dimensions like arousal and valence are more consistent with autonomic responses. Multiple autonomic measures may better capture discrete emotional states. Startle response magnitude is a measure of emotional valence, with larger responses to negative stimuli and smaller responses to positive ones. However, it does not assess discrete emotions. Startle responses are influenced by arousal and valence, and are more sensitive to high-arousal stimuli. Brain states, as measured by EEG and neuroimaging, show that left-hemisphere activation is associated with approach-related emotions, while right-hemisphere activation is associated with avoidance-related emotions. Neuroimaging studies suggest that specific brain regions (e.g., amygdala, insula, prefrontal cortex) are linked to different emotions, but findings are inconsistent and often confounded by induction methods. Behavioural measures, such as vocal characteristics and facial expressions, are sensitive to emotional valence. Vocal pitch and amplitude are associated with arousal, while facial expressions (e.g., Duchenne smiles) are linked to positive emotions. However, facial behaviours may not always reflect discrete emotions due to factors like culture and audience presence. Whole-body behaviour, such as posture, is associated with certain emotions (e.g., pride, embarrassment). These behaviours may be more relevant to social-status-related emotions. In conclusion, emotional responses are better understood through dimensional frameworks rather than discrete states. Different measures of emotion are sensitive to different dimensions, and no single measure is sufficient to capture the complexity of emotion. The review supports the idea that emotional states are best understood through a dimensional perspective, with multiple measures providing complementary insights.