Background: Child abuse appears to be on the increase during the COVID-19 pandemic, but the extent that lockdown measures modified the association between stress and abuses has not been systematically assessed .

Objectives: To assess : 1) the association between caregiver's stress and self-reported verbal abuse and corporal punishment of a child in the household, and; 2) modification of the stated association by experienced COVID-19 lockdown measures . Participants and settings: Caregivers residing in villages on lockdown in the Deep South of Thailand (n=466 participants)

Methods: We randomly sampled 12 villages in the study area, and 40 households per village . Trained enumerators who were residents of the sampled villages collected the data using phone-based interview . We measured stress level using the standard ST-5 questionnaire . We developed and pilot-tested questions for measurement of child abuse and lockdown experiences specifically for this study .

Results: Caregivers with moderate and higher levels of stress were more likely than caregivers with low level of stress to report verbal abuse (48% vs. 23%, respectively; Adj . OR = 3.12 , 95% CI = 1.89 , 5.15) and corporal punishment (28% vs. 8%, respectively; Adj . OR = 2.76 , 95% CI = 1.41 , 5.42). We found that COVID-19 lockdown experiences modified the associations between stress and verbal abuse and corporal punishment . Conclusion: There were associations between stress and abuses, which were modified by lockdown experiences . However, social desirability, lack of details in the answers, and potential confounding by mental illness co-morbidities were notable limitations of the study . Caveat is advised in the interpretation of the study findings.