Background: Digital droplet PCR (ddPCR) is a very sensitive high throughput genotyping methodology. To date, the use of ddPCR in immunohematology is restricted to fetal genotyping of red blood cell antigens. Our hypothesis is that this technology could be applied to screen for rare red blood cell genotypes, such as Di(b-). Methods: Nucleic acid of 3,168 donors was extracted for viral screening routine in pools of 6, which were converted into three types of 48-donor pools: control pools (only DI*B/B samples), pools with varying amount of DI*A/B samples (n=1 to 5) and a pool with one rare DI*A/A sample. Pools were genotyped using ddPCR to detect and quantify DI*A and DI*B alleles. Results: DI*A allele was accurately detected in all pools containing Di(a+b+) samples and in the pool containing one Di(a+b-) sample. No copies were detected in the control pools (n=60). The ratio between the number of DI*A and DI*B copies varied significantly between the pools and the triplicates. Conclusion: The proposed ddPCR assay was accurate in identifying the rare DI*A allele in large pools of donors and can be applied to screen for Di(b-) phenotype. The strategy can potentially be extended to search for other rare RBC phenotypes.