Mapping loss of heterozygosity (LOH) regions in the genomes of tumor tissues is a practical approach for identifying genes whose loss is related to tumorigenesis. Conventional LOH analyses using microsatellite or single nucleotide polymorphism (SNP) markers require the simultaneous examination of tumor- and matched normal-DNA. Here, we improved the previously developed SNP-based LOH assay using single strand conformation polymorphism (SSCP) analysis, so that LOH in tumor samples heavily contaminated with normal DNA can now be precisely estimated, even when matched normal DNA is not available. We demonstrate the reliability of the improved SSCP-based LOH detection method, called the LOH estimation by quantitative SSCP analysis using averaged control (LOQUS-AC), by comparing the results with those of the previous "LOH estimated by quantitative SSCP assay" (LOQUS) method. Using the LOQUS-AC assay, LOH was detected at a high consistency (98.1%) with the previous LOQUS method. We then applied this new method to characterize LOH profiles in 130 meningiomas, using 68 SNPs (i.e., a mean inter-SNP interval of 441 kbp) that are evenly distributed throughout chromosome 1p36. Benign, atypical and anaplastic meningiomas exhibited 1p36 LOH at frequencies of 48.39, 84.62 and 100.00%, respectively, using LOQUS-AC. Subsequently, we detected a candidate common LOH region on 1p36.11 that might harbor tumor suppressor genes related to malignant progression of meningioma.