Little is known about whether severe psychopathology influences the assessment of self-reported attachment style. Fifty-eight randomly selected adult psychiatric inpatients completed the Experiences in Close Relationship questionnaire (ECR; Brennan, Clark, & Shaver, 1998) and were administered the 24-item Brief Psychiatric Rating Scale (BPRS; Ventura et al., 1993) and the Hamilton Depression Rating Scale (HDRS; Hamilton, 1960) at both admission and discharge. The Structured Clinical Interview for DSM-IV (SCID-I; First, Spitzer, Gibbon, & Williams, 1996) was used to establish Axis I diagnoses. The ECR scales showed good internal consistency and absolute stability both in patients with (n = 24) and without (n = 34) a psychotic disorder. Relative stability was only fair among patients with psychotic disorders but good among patients without psychotic disorders. Neither higher BPRS or HDRS scores, nor the presence of a psychotic disorder, significantly reduced the retest reliability of the ECR scales. These findings suggest that self-report measures might provide a reliable assessment of attachment style in patients with severe psychopathology, except for the most severely impaired patients.