Objectives The purpose of this research was to create two state measures of self-compassion based on the Self-Compassion Scale (SCS): an 18-item State Self-Compassion Scale-Long form (SSCS-L) that could be used to measure the six components of self-compassion, and a six-item State Self-Compassion Scale-Short form (SSCS-S) that could be used as a measure of global state self-compassion. Methods Study 1 ( N = 588) used a community sample to select items for the SSCS-L and SSCS-S. Confirmatory Factor Analyses, Exploratory Structural Equation Modeling (ESEM), and bifactor modeling were used to analyze the factor structure of the SSCS-L and SSCS-S. Predictive validity was assessed by examining associations with positive and negative affect. Study 2 ( N = 411) used a student sample to examine the psychometric properties of the SSCS-L and SSCS-S after a self-compassion mindstate induction (SCMI) to determine if its factor structure would remain unchanged after manipulation. Study 3 ( N = 139) examined the psychometric properties of the SSCS-S alone. Results The SSCS-L had good psychometric properties and SSCS-S was also adequate. A bifactor-ESEM representation (with one global factor and six components) was supported for the SSCS-L, and a single factor was supported for the SSCS-S. Both scales were reliable. Psychometric properties were unchanged after the experimental manipulation of self-compassion. A total state self-compassion score and subscale scores were associated with positive and negative affect in the expected directions. Conclusions The SSCS-L and SSCS-S appear to be valid measures of state self-compassion.