Some monogenic disorders are characterized by a vast genetic heterogeneity. In individuals with similar clinical phenotype, causative mutations can be found in one gene from a subset described as implicated in the disease. Such genetic heterogeneity limits considerably the diagnostic offer for the patients, and a majority is left without molecular diagnosis. We developed an alternative diagnostic approach by targeted high throughput sequencing (specific to the coding regions of genes of interest by a technique of exon capture) through three genetically heterogeneous disorders: Bardet-Biedl syndrome (19 genes reported), leukodystrophies (50 genes), and intellectual disability (>400 genes). In light of its efficiency, this approach has since been implemented in diagnostic routine for Bardet-Biedl syndrome and intellectual disability (80% and 25% of diagnostic yields respectively, significantly higher than those of previous methods). Beyond diagnosis, this approach allows unbiased means to assess the contribution of each gene in the disease and highlight recurrent genes, and establish new correlations genotype to phenotype, overall providing much insight in the genetics of a particular disease.