The Deciphering Developmental Disorders (DDD) study aimed to identify the prevalence and architecture of *de novo* mutations (DNMs) in individuals with severe, undiagnosed developmental disorders (DDs). The study exome sequenced 4,293 families and meta-analysed data from an additional 3,287 individuals with similar disorders. Key findings include: 1. **Sex and Family Relatedness**: Males had a lower chance of carrying likely pathogenic DNM compared to females. Individuals with affected relatives and those from consanguineous unions were less likely to have pathogenic DNMs. 2. **Paternal Age**: Paternal age was a significant factor influencing the number of DNMs, with a strong effect on the likelihood of having a pathogenic DNM. 3. **Genome-Wide Significance**: 93 genes were significantly enriched for damaging DNMs, including 14 without previous compelling evidence. These genes were characterized for phenotypic diversity. 4. **Prevalence of Pathogenic DNMs**: 42% of the cohort carried pathogenic DNMs in coding sequences, with approximately half disrupting gene function and the remainder altering function. 5. **Birth Prevalence**: Developmental disorders caused by DNMs have an average birth prevalence of 1 in 213 to 1 in 448, depending on parental age, equating to nearly 400,000 children born per year. 6. **Gene Discovery**: The study identified 47% of haploinsufficient DD-associated genes previously robustly associated with neurodevelopmental disorders. 7. **Mechanisms of Action**: The relative proportion of altered-function and loss-of-function mechanisms among excess DNMs was estimated, with 57% likely acting by loss-of-function and 43% by altered-function. 8. **Power of Sequencing Strategies**: Exome sequencing was found to be more powerful for novel disease gene discovery compared to genome sequencing, given current cost and sensitivity considerations. The study highlights the significant role of *de novo* mutations in severe developmental disorders and underscores the importance of integrating phenotypic data for improved gene discovery and understanding of disease mechanisms.The Deciphering Developmental Disorders (DDD) study aimed to identify the prevalence and architecture of *de novo* mutations (DNMs) in individuals with severe, undiagnosed developmental disorders (DDs). The study exome sequenced 4,293 families and meta-analysed data from an additional 3,287 individuals with similar disorders. Key findings include: 1. **Sex and Family Relatedness**: Males had a lower chance of carrying likely pathogenic DNM compared to females. Individuals with affected relatives and those from consanguineous unions were less likely to have pathogenic DNMs. 2. **Paternal Age**: Paternal age was a significant factor influencing the number of DNMs, with a strong effect on the likelihood of having a pathogenic DNM. 3. **Genome-Wide Significance**: 93 genes were significantly enriched for damaging DNMs, including 14 without previous compelling evidence. These genes were characterized for phenotypic diversity. 4. **Prevalence of Pathogenic DNMs**: 42% of the cohort carried pathogenic DNMs in coding sequences, with approximately half disrupting gene function and the remainder altering function. 5. **Birth Prevalence**: Developmental disorders caused by DNMs have an average birth prevalence of 1 in 213 to 1 in 448, depending on parental age, equating to nearly 400,000 children born per year. 6. **Gene Discovery**: The study identified 47% of haploinsufficient DD-associated genes previously robustly associated with neurodevelopmental disorders. 7. **Mechanisms of Action**: The relative proportion of altered-function and loss-of-function mechanisms among excess DNMs was estimated, with 57% likely acting by loss-of-function and 43% by altered-function. 8. **Power of Sequencing Strategies**: Exome sequencing was found to be more powerful for novel disease gene discovery compared to genome sequencing, given current cost and sensitivity considerations. The study highlights the significant role of *de novo* mutations in severe developmental disorders and underscores the importance of integrating phenotypic data for improved gene discovery and understanding of disease mechanisms.