Copy number variation （CNV） is a type of genetic variation which may have important roles in phenotypic variability and disease susceptibility. To hunt for genetic variants underlying human height variation, we performed a genome wide CNV association study for human height in 618 Chinese unrelated subjects using Affymetrix 500K array set. After adjusting for age and sex, we found that four CNVs at 6p21.3, 8p23.3-23.2, 9p23 and 16p12.1 were associated with human height （with borderline significant p value： 0.013, 0.011, 0.024, 0.049; respectively）. However, after multiple tests correction, none of them was associated with human height. We observed that the gain of copy number （more than 2 copies） at 8p23.3-23.2 was associated with lower height （normal copy number vs. gain of copy number： 161.2 cm vs. 153.7 cm, p = 0.011）, which accounted for 0.9% of height variation. Loss of copy number （less than 2 copies） at 6p21.3 was associated with 0.8% lower height （loss of copy number vs. normal copy number： 154.5 cm vs. 161.1 cm, p = 0.013）. Since no important genes influencing height located in CNVs at loci of 8p23.3-23.2 and 6p21.3, the two CNVs may cause the structural rear- rangements of neighbored important candidate genes, thus regulates the variation of height. Our results expand our knowledge of the genetic factors underlying height variation and the biological regulation of human height.
