Noninvasive prenatal testing (NIPT), which involves analysis of circulating cell-free fetal DNA (cffDNA) from maternal plasma, is highly effective for detecting feto-placental chromosome aneuploidy. However, recent studies suggested that coverage-based shallow-depth NIPT cannot accurately detect smaller single or multi-loci genetic variants. To assess the fetal genotype of any locus using maternal plasma, we developed a novel genotyping algorithm named pseudo tetraploid genotyping (PTG). We performed paired-end captured sequencing of the plasma cell-free DNA (cfDNA), in which case a phenotypically healthy woman is suspected to be carrying a fetus with genetic defect. After a series of independent filtering of 111,407 SNPs, we found one variant in COL1A1 graded with high pathogenic potential which might cause osteogenesis imperfecta (OI). Then, we verified this mutation by Sanger sequencing of fetal and parental blood cells. In addition, we evaluated the accuracy and detection rate of the PTG algorithm through direct sequencing of the genomic DNA from maternal and fetal blood cells. Collectively, our study developed an intuitive and cost-effective method for the noninvasive detection of pathogenic mutations, and successfully identified a de novo variant in COL1A1 (c.2596 G > A, p.Gly866Ser) in the fetus implicated in OI.