This Registered Report provides the first test of measurement invariance across time points and estimates of test-retest reliability for the Social Thermoregulation, Risk Avoidance Questionnaire (STRAQ-1, Vergara et al., 2019). The scale was developed and validated to understand the physiological drives underlying interpersonal bonding, measured by four constructs: the desire to socially regulate one’s temperature, the desire to solitary regulate one’s temperature, the sensitivity to higher temperatures, and the desire to avoid risk. Previous studies with large samples across 12 countries showed that the STRAQ-1 has a stable factorial structure, satisfying internal consistencies for the temperature subscales, and expected correlations in its nomological network. However, to date, this instrument has no estimates of test-retest reliability. Throughout four academic years (from 2018 to 2022), N = 183 French student participants took the STRAQ-1 at least two times. Out of the four STRAQ-1 subscales, only two were longitudinally invariant across two-time points. Thus, half of the STRAQ-1 constructs and latent scores were dissimilar and incomparable across time. We then conducted test-retest reliability using Intra Class Correlation coefficient (ICC) for the Social Thermoregulation, Solitary Thermoregulation, High-Temperature Sensitivity, and Risk Avoidance subscales. ICCs estimates were respectively for agreement and consistency: .70, .70 overall moderate to good, .62, .62 overall moderate, .67, .67 overall moderate, and .53, .53 overall poor to moderate, respectively. Our study suggests that test-retest reliability was insufficient for psychological diagnosis, and that future studies should address the problem of low generalizability of the constructs.