PurposeTo identify genetic etiologies and genotype/phenotype associations for unsolved ocular congenital cranial dysinnervation disorders (oCCDDs). MethodsWe coupled phenotyping with exome or genome sequencing of 467 probands (550 affected and 1108 total individuals) with genetically unsolved oCCDDs, integrating analyses of pedigrees, human and animal model phenotypes, and de novo variants to identify rare candidate single-nucleotide variants, insertion/deletions, and structural variants disrupting protein-coding regions. Prioritized variants were classified for pathogenicity and evaluated for genotype/phenotype correlations. ResultsAnalyses elucidated phenotypic subgroups, identified pathogenic/likely pathogenic variant(s) in 43 of 467 probands (9.2%), and prioritized variants of uncertain significance in 70 of 467 additional probands (15.0%). These included known and novel variants in established oCCDD genes, genes associated with syndromes that sometimes include oCCDDs (eg, MYH10, KIF21B, TGFBR2, and TUBB6), genes that fit the syndromic component of the phenotype but had no prior oCCDD association (eg, CDK13, TGFB2), genes with no reported association with oCCDDs or the syndromic phenotypes (eg, TUBA4A, KIF5C, CTNNA1, KLB, and FGF21), and genes associated with oCCDD phenocopies that had resulted in misdiagnoses. ConclusionThis study suggests that unsolved oCCDDs are clinically and genetically heterogeneous disorders often overlapping other Mendelian conditions and nominates many candidates for future replication and functional studies.