In response to the coronavirus disease (COVID-19) pandemic, governments have adopted a variety of public health measures. In this study, we aimed to evaluate the impact of testing on the fatality rate. We use data on inpatients across French geographic areas and propose a novel methodology that exploits policy discontinuities at region borders to estimate the effect of testing symptomatic individuals on the case-fatality rate in France. Our identification strategy is based on the fact that, in France, testing policies are determined regionally by the Regional Public Health Agencies. We compare all contiguous department pairs located on the opposite sides of a region border. Department pairs have different testing rates but share similar health trends. The heterogeneity in testing rate between department pairs together with the similarities in other dimensions allow us to mimic the existence of treatment and control groups and to identify the impact of testing on the mortality rate. We find that in France, the increase of one percentage point in the test rate is associated with a decrease of 0.001 percentage point in the death rate. Putting this number into perspective involves that for each additional 1000 tests, one person would have remained alive.