In this paper, we evaluate the impact of various factors on the propensity of witnesses to report Intimate Partner Violence (IPV). In particular, we analyze whether requiring the identification of the witness and/or the aggressor can be barriers to reporting IPV. In addition, we also explore whether different institutional channels can also have an impact on the propensity to report IPV. To perform this evaluation, we designed an experiment with 969 individuals (representative sample of the population of Navarra between 18 and 48) that were placed in the role of a witness in a situation of IPV. We use an incentivized vignette and four different treatments to respond to our research questions. Our findings indicate that having to identify the aggressor reduces the propensity to report IPV by 7.3 percentage points while the requirement of identifying oneself reduces the propensity to report IPV by 4.8 percentage points. In addition, the propensity to report IPV via police instead of via social services, whenever there is the double requirement of identifying oneself and the aggressor, is 8.2 percentage points higher. We discuss the policy implications of these findings and propose alternatives to overcome these barriers.