Microhaplotypes (MHs), composed of two or more single nucleotide polymorphisms (SNPs) within short regions, have emerged as powerful markers for forensic applications due to advances in massively parallel sequencing (MPS). Recently, nanopore sequencing has surfaced as a third-generation long-read platform offering cost-effectiveness, real-time sequencing, and portability. Nevertheless, nanopore sequencing has a relatively high error rate that may reduce analytical reliability. Several studies have applied nanopore sequencing to forensic markers, yet noise levels related to reliability have not been systematically evaluated. In this study, we established a PCR-based short-amplicon (<290 bp) nanopore sequencing method for an in-house MH panel, and evaluated its performance by assessing genotype concordance rate and noise levels against previously generated Illumina MiSeq data. Target enrichment and sample barcoding were performed by a two-step PCR method using a PCR Barcoding Expansion kit. Adapter-ligated nanopore libraries were sequenced on an ONT MinION platform, with an R10.4.1 flow cell. MPS data for 33 MH markers were analyzed using STRait Razor 3.0 and Visual Microhap. Marker coverage balance was about 2.6-fold, ranging from 7,961 to 20,714. Genotype concordance rate between nanopore and Illumina platforms was 99.86%, with three genotype discordances. Two discordances at mh11KK-183 were attributed to homopolymer errors, and one at mh16KK-011 resulted from allele imbalance. Noise levels were 0.14% for Illumina MiSeq and 6.21% for ONT MinION with a Background noise level set to 1%. In conclusion, this study demonstrated the applicability of nanopore sequencing to MH short-read analysis. However, further noise mitigation will be required to apply this approach to forensic casework samples exhibiting high noise levels.