Polymerase chain reaction (PCR) has proven to be the gold-standard for SARS-CoV-2 detection in clinical settings. The most common approaches rely on nasopharyngeal specimens obtained from swabs, followed by RNA extraction, reverse transcription and quantitative PCR. Although swab-based PCR is sensitive, swabbing is invasive and unpleasant to administer, reducing patient compliance for regular testing and resulting in an increased risk of improper sample collection. To overcome these obstacles, we developed a non-invasive one-step RT-qPCR assay performed directly on saliva specimens. The University of Nottingham Asymptomatic Testing Service (UoNATS) protocol simplifies sample collection and bypasses the need for RNA extraction, additives, or extraneous processing steps, thus helping to encourage more regular testing and reducing processing time and costs. We have evaluated the assay against the performance criteria specified by the UK regulatory bodies and attained accreditation (BS EN ISO/IEC 17025:2017) for SARS-CoV-2 diagnostic testing by the United Kingdom Accreditation Service (UKAS). We observed a sensitivity of 1 viral copy per microlitre of saliva and demonstrated a concordance of >99.4% between our results and those of other accredited testing facilities. We concluded that saliva is a stable medium with surprising longevity, and allows for a highly precise, repeatable, and robust testing method.