The Self-compassion Scale-Short Form (SCS-SF) was developed as an economical alternative for the Self-compassion Scale (SCS), one of the few scales to assess self-compassion. Despite the active use of the SCS-SF, a psychometric evaluation of this scale remains limited. This study analysed the factor structure, reliability, and construct validity of the SCS-SF in UK university student populations. Of 365 students approached, 333 completed the SCS-SF and other measures including negative psychological constructs (mental health problems, self-criticism, and mental health shame) and positive psychological constructs (selfreassurance, resilience, and well-being). Data were analysed through confirmatory factor analyses and correlations. CFA revealed that the six-factor structure, reported in the validation paper, was not replicated. The positive factor, consisting of the three positive subscales, was not strongly related to any variable but moderately related to reassured-self, resilience, well-being, and inadequate-self. The negative factor, consisting of the three negative subscales, was strongly related to inadequate-self and moderately related to resilience, reassured-self, stress, well-being, depression, and internal shame. Coefficients in the negative factor were in general larger than those in the positive factor. The total SCS-SF score was most strongly related to inadequate-self, followed by resilience. Inter-correlations of the six subscales did follow neither Neff's [1] theoretical model of self-compassion nor the full-scale factor solution. Findings do not accord with the common use of the global SCS-SF score as an assessment of six factors of self-compassion and suggest a two-factor solution assessing self-criticism and self-compassion.