Aim The aim of the present study is to further the understanding of who cries in therapy and the relation of technique with crying behaviour in therapy. Method Psychological assessment feedback sessions, prior to the initiation of formal therapy for 52 patients beginning psychotherapy at a university‐based clinic were coded for discrete crying segments. Data about patient characteristics and the process of the session were collected at the time of the session. Therapist's interventions were recorded verbatim and independently rated. Results The number of times a patient cried during their session correlated negatively with global assessment of functioning scores and positively with measures of borderline personality disorder pathology as well as a measure of severity of childhood sexual abuse. Patients' crying behaviour demonstrated significant negative correlations with the overall experience of the session (bad/good), smoothness and positivity. Group differences between criers and non‐criers reflected these trends as well. No significant correlations or group differences were found with regard to patient‐rated or therapist‐rated alliance as it relates to crying behaviour. Analysis indicates that therapist intervention prior to patient crying most often encouraged the exploration and expression of difficult affect, new perspectives on key issues or the patient's fantasies and wishes. Discussion Our study addresses a significant gap in the clinical literature on crying. Crying behaviour seems to be related to certain clinical variables and has a negative impact on patient experience of the session in which they cry, although the alliance remained unaffected. Limitations Small sample, outpatients with mild/moderate psychopathology and graduate trainees provided therapy. Copyright © 2013 John Wiley & Sons, Ltd. Key Practitioner Message Patients with greater problems in emotional dysregulation, borderline personality disorder symptoms and greater severity of childhood sexual abuse are more likely to display greater affective intensity during the beginning of treatment. Results suggest that the alliance may remain strong despite patients experiencing a session in which they cried as difficult. Therapeutic interventions that focus on affect, new understanding of old patterns and patient fantasies with outpatient clinical populations appeared to be associated with crying in session.