BACKGROUND: An estimated 1.8 billion children live in countries where COVID-19 disrupted violence prevention and response. It is important to understand how government policies to contain COVID-19 impacted children's ability to seek help, especially in contexts where there was limited formal help-seeking prior to the pandemic. We aimed to quantify how the national lockdown in Zimbabwe affected helpline calls for violence against children, estimated the number of calls that would have been received had the lockdown not occurred and described characteristics of types of calls and callers before and after the national lockdown. METHODS: We used an interrupted time series design to analyse the proportion of violence related calls (17,913 calls out of 57,050) to Childline Zimbabwe's national child helpline between 2017 to 2021. We applied autoregressive integrated moving average regression (ARIMA) models to test possible changes in call trends before and after the March 2020 lockdown and forecasted how many calls would have been received in the absence of lockdown. In addition, we examined call characteristics before and after lockdown descriptively. RESULTS: The proportion of violence related calls decreased in the 90 days after the lockdown and subsequently returned to pre-COVID-19 levels. We estimate that 10.3% (95% confidence interval [CI] 6.0-14.6%) more violence related calls would have occurred in this period had there not been a lockdown. Violence was increasingly reported as occurring in children's households, with fewer reports from children and formal child protection actors. CONCLUSIONS: Lockdowns dramatically change everyday life and strain populations, which is unlikely to reduce violence prevalence but may reduce help-seeking. The three months after COVID-19 lockdowns may be key time periods when help-seeking for violence decreases drastically. Policy makers should ensure that in-person and remote services support help-seeking. Interventions and campaigns may additionally want to target adult female family members in encouraging reporting of suspected violence cases when they occur within households and are perpetuated by other family members. We suggest a composite approach of scaling-up remote reporting mechanisms that are accessible and geographically well-distributed, establishing non-traditional sites for help seeking within communities and continuing limited in-person home visitation for known cases of violence.