Expressed emotion (EE), is a measure of the family environment that is based on how the relatives of a psychiatric patient spontaneously talk about the patient. It specifically measures three to five aspects of the family environment: the most important are critical comments, hostility, emotional over-involvement, with positivity and warmth sometimes also included as indications of a low-EE environment. The psychiatric measure of expressed emotion is distinct from the general notion of communicating emotion in interpersonal relationships, and from another psychological metric known as family emotional expressiveness.

A high level of EE in the home can worsen the prognosis in patients with mental illness, such as schizophrenia and social anxiety disorder, or act as a potential risk factor for the development of psychiatric disease. Higher degrees of expressed emotion in the environment of a patient have been empirically found to be robust predictors of relapse of schizophrenia, eating disorder, and mood disorders. It has also been investigated as a contributor to the progress of unipolar depression, bipolar disorder, dementia, and diabetes. Interventions to improve outcomes include reducing contact with high-EE caregivers, and educating and supporting families so they can reduce high-EE behavior.

Various mechanisms have been proposed to explain why high EE family environments produce worse outcomes, including that:

Typically it is determined whether a person or family has high EE or low EE through a taped interview known as the Camberwell Family Interview (CFI). Answers to questions and non-verbal cues are used to determine if someone has high expressed emotion. There is another measurement that is taken from the view of the patient, which rates the patient's perception of how their family feels about them and the disorder. An alternative measure of expressed emotion is the Five Minutes Speech Sample (FMSS), where the relatives are asked to talk about the patient for five uninterrupted minutes. Although this measure requires more training, it becomes a quicker form of assessment than the CFI.

A 1956 study of readmissions of schizophrenia patients in London by George Brown found that patients discharged to live with their parents or wives were more frequently readmitted than those discharged to live with siblings or non-family in lodging houses. It also found that those that lived with their mothers were more likely to be readmitted if the mothers did not work outside the home, suggesting that the duration of exposure to certain family members was related to relapse. Brown devised the five dimensions of expressed emotion to quantify the interpersonal environmental exposures of patients.

The advantage of a low-EE environment has been cited to partly explain the success of the Belgian village of Geel, where residents have for hundreds of years welcomed unrelated people with mental illness to live with them.

Janis H. Jenkins and her team conducted the first study showing that Mexican immigrants’ familial emotional responses of warmth and sympathy toward mentally ill kin in the United States contributes to a more favorable course of illness than in the case of their Euro-American counterparts.

Family members with high expressed emotion are hostile, very critical and not tolerant of the patient. They feel like they are helping by having this attitude. They not only criticize behaviors relating to the disorder but also other behaviors that are unique to the personality of the patient. High expressed emotion is more likely to cause a relapse than low expressed emotion.