Pregnancy loss (PL), particularly early pregnancy loss (EPL), is a prevalent reproductive complication, with approximately 15% of confirmed pregnancies affected. Chromosomal abnormalities are implicated in more than half of EPLs, with trisomies being the most prevalent. Partial abnormalities, including segmental deletions, duplications, and unbalanced translocations, are detected in up to 10% of EPL cases. This study focuses on the precise characterization of partial chromosomal abnormalities, previously identified by Quantitative fluorescent polymerase chain reaction (QF-PCR) and multiplex ligation probe amplification (MLPA) analyses. By employing an array comparative genomic hybridization (aCGH), we analyzed 20 EPL samples, identifying 32 partial chromosomal abnormalities, including 18 deletions and 14 duplications, with an average size of 33.2 Mb. Notably, two abnormalities previously undetected by QF-PCR and MLPA were revealed (deletions in 7q36, and 1p36.32p36.31regions), emphasizing the necessity of high-resolution genomic tools. Chromosomes 1, 18, and 13 emerged as frequently involved, aligning with previous associations with recurrent pregnancy loss. Recurrent abnormalities were identified in six chromosomal regions, with chromosome 1p36.33-p36.32 exhibiting the highest frequency. Gene Ontology (GO) enrichment analysis of recurrent regions highlighted disruptions in critical biological processes, including molecular binding, enzymatic activity, and cellular development. Many genes in these regions are linked to multisystem syndromes, suggesting their involvement in early embryonic development and pregnancy viability. Our findings underscore the complexity of EPL's genetic landscape, demonstrating that large CNVs, may disrupt multiple genes critical for development. Although, subtelo-meric MLPA reliably detects telomeric partial chromosomal abnormalities in EPLs, aCGH is essential for detection and precise characterization of all CNVs, thus enhancing diagnostic and counseling strategies in affected couples.