Congenital cytomegalovirus (cCMV) is the most common perinatal infection, the leading cause of nongenetic sensorineural hearing loss, and one of the leading causes of neurodevelopmental impairment in the developed world. Early identification via newborn screening (NBS) would benefit the many undiagnosed infants who are either asymptomatic or mildly to moderately symptomatic, of whom 20% develop sequelae. The sensitivity of a recently developed PCR-based method to detect CMV in dried blood spots (DBS) is less than 80% and requires significantly more specimen than any other NBS test. We sought to improve the analytical sensitivity of the screening method by using droplet digital PCR and direct PCR and decreasing the amount of specimen utilized. The methods were tested with CMV-spiked filters, DBS from CMV-spiked cord blood, and DBS from neonates with cCMV. The results showed that the analytical sensitivity of all modified methods was equivalent to that of the reference method, with consistent CMV detection at high viral loads and inconsistent detection at low viral loads. IMPORTANCE Implementation of screening for cCMV in public health programs is hindered by feasibility challenges, including limited specimen availability and an insufficiently sensitive DBS-based screening assay. We report on efforts to improve the currently available DBS-based molecular assay to increase its feasibility of implementation in newborn screening programs. Although the analytical sensitivity of the modified methods was similar at the lower IU, equivalent CMV detection was achieved using one punch instead of the required three punches for the reference method. This reduction in sample size has the potential to substantially improve feasibility of NBS for cCMV. A population-based study is needed to further evaluate the clinical sensitivity of the improved assay.