Proper development of the vertebrate embryo requires fusion of epithelial cell sheets, resulting in closure of the developing neural tube, palate and retina. Within the eye, failure of the ocular fissure to close results in ocular coloboma. Together with the etiologically related disorders microphthalmia (small eyes) and anophthalmia (no eyes), coloboma represents up to 11% of all pediatric blindness. Previous work from our laboratory and others has defined a key role for Bone Morphogenetic Protein (BMP) signaling within the retina in regulating ocular fissure closure. Outside the retina, a population of neural crest cells known as periocular mesenchyme (POM) migrates to the ocular fissure sites and is required for fissure closure, although our understanding of the mechanism remains limited. Here we identify a novel regulator of ocular fissure closure: bmp3. Exome and Sanger sequencing of 480 patients with coloboma has identified five with variants in BMP3, three of which are in the mature domain and predicted to damage protein function. The high degree of conservation at these residues, and absence of such variants in exome databases and control samples, demonstrates the likely pathogenicity of these mutations. Studies in cultured cells demonstrate that the identified BMP3 variants have altered activity in vitro. In zebrafish, morpholino inhibition of Bmp3 results in small eyes and fissure closure defects, implicating this ligand as a key regulator of eye morphogenesis. bmp3 is an intriguing candidate as it is not expressed within the eye, but in cells immediately anterior to the eye, unlike any previously identified BMP ligand. Intriguingly, knockdown of Bmp3 activity causes defects in cranial neural crest cell development and fewer POM cells surrounding the eye. We hypothesize that bmp3 is a novel factor that governs neural crest cell development. Consistent with this, bmp3 has been previously implicated in craniofacial development. To test this hypothesis in a bmp3 null, we have used the CRISPR-Cas9 system to generate a frameshift mutation that results in a premature stop codon upstream of the mature domain of Bmp3. Current studies are focused on defining neural crest and ocular phenotypes in a bmp3 mutant background and elucidating the molecular mechanisms underlying Bmp3 function in neural crest development.