Background: SMN1 and SMN2 are causative and modifier genes, respectively, for spinal muscular atrophy (SMA). The incidence of SMN1 homozygous deletion in Japan is 1 in 20,000. However, the incidence of SMN2 homozygous deletion in Japan remains unknown. Methods: To clarify the incidence of homozygous SMN2 deletion in Japan, real-time polymerase chain reaction (PCR) was performed on dried blood spot (DBS) samples collected from newborns nationwide. Samples with positive or ambiguous results were retested using PCR-restriction fragment length polymorphism (PCR-RFLP) and nucleotide sequence analysis. Results: Of the 1000 DBS samples that were screened using real-time PCR, 51 were positive. Retesting using PCR-RFLP analysis identified 10 false results: six false positives and four false negatives. Therefore, there were 49 true positives among the 1000 samples. Notably, nucleotide sequence analysis revealed that the false negatives were caused by the cross-reactivity of SMN2 primers with SMN1 sequences. Conclusions: The incidence of homozygous SMN2 deletion in Japan is approximately 1 in 20 people. This incidence is much higher than that of homozygous SMN1 deletion and may reflect the vulnerability of the SMN2 region. Importantly, the results of the present study suggest that false negatives in the screening process were caused by cross-reactivity with non-target gene sequences.