Abstract Background Testing for viral DNA in neonatal blood dried on paper (DBS) has proved a valid means of diagnosing congenital CMV infection with both clinical and epidemiological relevance. To assess the quality of the detection of CMV-DNA on DBS in laboratories performing this test a proficiency panel consisting of nine samples with two blood spots on each filter paper was produced and distributed. Six samples were derived from whole blood, negative for CMV DNA and antibody, and spiked with cell-grown CMV Towne in various concentrations (7.3 × 10 2 – 9.6 × 10 5 copies/ml), one was a CMV positive clinical specimen (3.9 × 10 6 copies/ml), and two samples were CMV-negative whole blood. Results The 27 responding laboratories from 14 countries submitted 33 datasets obtained by means of conventional PCR (n = 5) or real-time PCR (n = 28) technologies. A correct positive result was reported in at least 91% of datasets in samples with a viral load of 8.8 × 10 4 copies/ml or higher. However only 59% and 12% identified the 9.4 × 10 3 and 7.3 × 10 2 copies/ml samples, respectively, correctly as positive. False positive results were reported by 9% of laboratories and in 11% of datasets. Conclusion These results indicate a clear need for improvement of methods as sensitivity and false-positivity still appear to be a major problem in a considerable number of laboratories.