Pseudoexfoliation syndrome (PEXS) is an age-related elastosis, strongly associated with the development of secondary glaucoma. It is clearly suggested that PEXS has a genetic component, but this has not been extensively studied. Here, a genome-wide association study (GWAS) using a DNA-pooling approach was conducted to explore the potential association of genetic variants with PEXS in a Polish population, including 103 PEXS patients without glaucoma and 106 perfectly (age- and gender-) matched controls. Individual sample TaqMan genotyping was used to validate GWAS-selected single-nucleotide polymorphism (SNP) associations. Multivariate binary logistic regression analysis was applied to develop a prediction model for PEXS. In total, 15 SNPs representing independent PEXS susceptibility loci were selected for further validation in individual samples. For 14 of these variants, significant differences in the allele and genotype frequencies between cases and controls were identified, of which 12 remained significant after Benjamini-Hochberg adjustment. The minor allele of five SNPs was associated with an increased risk of PEXS development, while for nine SNPs, it showed a protective effect. Beyond the known LOXL1 variant rs2165241, nine other SNPs were located within gene regions, including in OR11L1, CD80, TNIK, CADM2, SORBS2, RNF180, FGF14, FMN1, and RBFOX1 genes. None of these associations with PEXS has previously been reported. Selected SNPs were found to explain nearly 69% of the total risk of PEXS development. The overall risk prediction accuracy for PEXS, expressed by the area under the ROC curve (AUC) value, increased by 0.218, from 0.672 for LOXL1 rs2165241 alone to 0.89 when seven additional SNPs were included in the proposed 8-SNP prediction model. In conclusion, several new susceptibility loci for PEXS without glaucoma suggested that neuronal development and actin remodeling are potentially involved in either PEXS onset or inhibition or delay of its conversion to glaucoma.