Background: Classical psychopathology greatly valued the interaction between clinician and patient, and assigned to the clinician's subjective experience a significant role in the diagnostic process. Psychoanalysis, too, ascribed a privileged position to the clinician's feelings and empathic participation in the assessment and deep understanding of the patient. This study aimed at testing the traditional, though still relatively unexplored empirically, tenet that particular diagnostic groups elicit distinct and diagnostically useful reactions from clinicians. Sampling and Methods: The study was performed in several psychiatric inpatient and outpatient units in Rome, Italy. The clinicians completed the Assessment of Clinician's Subjective Experience (ACSE) questionnaire and other standardized assessment instruments when they evaluated a previously unknown patient. All adult patients diagnosed with schizophrenia (n = 119), cluster B personality disorder (n = 114), manic or mixed bipolar I episode (n = 59), and unipolar depression or anxiety disorder (n = 130) were included in the study, for a total of 422 patients evaluated by 35 clinicians. Results: We found a significant and theoretically consistent relationship between the clinicians' pattern of subjective experience during the first visit and patients' clinical diagnoses. Patients with unipolar depression/anxiety showed significantly lower scores than the other groups on all ACSE scales except engagement; patients with schizophrenia displayed significantly higher scores than the other groups on difficulty in attunement, and significantly higher scores than patients with cluster B personality disorder on impotence. Compared with the other groups, the patients with cluster B personality disorder displayed significantly lower scores on engagement, and significantly higher scores on disconfirmation. In multivariate models controlling for patient's age and education, symptom severity, clinician's sex, duration of visit and setting, diagnosis remained a significant predictor of scores on all ACSE scales except for impotence. Conclusions: The main limitations of the study are its reliance on clinical diagnoses and the non-independence of assessments. Further studies based on diagnoses made by a third observer through standardized instruments are needed to provide a most stringent test of the hypothesis that different diagnoses are associated with distinct profiles of clinicians' subjective experience. This study provided intriguing, though preliminary, evidence that the clinician's subjective experience may play a useful role in the diagnostic process. Time may have come to reintroduce the concept of intersubjectivity at the core of the diagnostic process.