Objective: Although the self-report version of Liebowitz Social Anxiety Scale (LSAS) is frequently used to measure social anxiety, data is lacking on the smallest detectable change (SDC), an important index of measurement error. We therefore aimed to determine the SDC of LSAS.

Methods: Japanese adults aged 20–69 years were invited from a panel managed by a nationwide internet research agency. We then conducted a test–retest internet survey with a two-week interval to estimate the SDC at the individual (SDC_ind) and group (SDC_group) levels.

Results: The analysis included 1300 participants. The SDC_ind and SDC_group for the total fear subscale (scoring range: 0–72) were 23.52 points (32.7%) and 0.65 points (0.9%), respectively. The SDC_ind and SDC_group for the total avoidance subscale (scoring range: 0–72) were 32.43 points (45.0%) and 0.90 points (1.2%), respectively. The SDC_ind and SDC_group for the overall total score (scoring range: 0–144) were 45.90 points (31.9%) and 1.27 points (0.9%), respectively.

Conclusions: Measurement error is large and indicate the potential for major problems when attempting to use the LSAS to detect changes at the individual level. These results should be considered when using the LSAS as measures of treatment change.