Interest has been growing in the mental health benefits of self-compassion. Whereas, most research on this topic has been conducted with the 26-item Self-Compassion Scale (SCS), a briefer 12-item version of the instrument, the Self-Compassion Scale-Short Form (SCS-SF), also exists. The SCS-SF has demonstrated good validity and reliability in non-clinical samples, but it has not been used often in research with psychotherapy clients. This study was designed to examine the factor structure and construct validity of the SCS-SF in a clinical population. Data for this study were collected from 1609 college students receiving services at 10 campus counseling centers. The previously proposed factor structure of the SCS-SF was not supported. Instead, analyses revealed two factors, Self Care and Self Disparagement. Evidence for the construct validity of these factors was found via expected relationships with indices of depression, anxiety, social anxiety, hostility, academic distress, eating concerns, family distress, maladaptive perfectionism, suicidality, self-injurious behavior, and social support. SCS-SF scores were unrelated to various measures of substance use. Implications for clinical work and future research are discussed.