Forensic perpetrator identification is a crucial process depending on accurate processing of the available evidence. In many cases where perpetrators are captured via CCTV footage of a crime, this material is challenging to process - typically due to low quality and limited information available. Over the past decade, the media has suggested that so-called Super-Recognizers (SRs), individuals with superior ability for processing facial identity, could improve perpetrator identification for law enforcement. In reality, such reports are anecdotal success stories - a formal Test of SRs’ ability has never been conducted. Using authentic forensic case material, we performed the first-ever comparison between SRs and control participants to answer the question: Can SRs improve perpetrator identification accuracy?