The development of Insensitive Munitions (IM) has been progressing for over three decades. Ever since the highly publicized US Navy aircraft carrier accidents in the 1960s to 1980s, there has been a growing recognition of the value of IM. Reminders of the need for IM have been provided all too often in the form of accidents, such as experienced by the US Army at Camp Doha and the prevalence of attacks on military installations around the world. The process for developing IM has improved over the years as technology for mitigating the consequences of accidental initiation has emerged. Early IM developments were based upon replacement of the traditional TNT-based explosives, with their high vulnerability, with reduced vulnerability PBXs. This led to significant improvements, such as that observed with the replacement of H-6 with PBXN-109 in the US Navy Mk82 GP bomb. From the early 1990s, the use of a complete systems approach was highlighted as the optimum method to achieve IM compliance while maintaining or enhancing operational performance. The use of a systems approach has resulted in the fielding of a number of munition systems with significant IM properties. The challenge for the future is to continue the development and fielding of improved performance IM munitions with limited funding for research and the high cost of introducing new ingredients into energetic formulations. A key development to allow continued progress to occur is the introduction of improved versions of current explosive ingredients. The attention focused in the past few years on forms of RDX with reduced shock sensitivity has highlighted the possibility of improving well-known materials. In the near future, the application of materials technology may provide improved versions of other important crystalline energetic materials currently in production or advanced development, including HMX, NTO, CL-20 and ADN and help advance the development of further explosive ingredients such as FOX-7 and LLM-105. Advances in the development and application of computer modeling must be made if we are to move forward from our current reliance on a limited number of canonical tests that are held to be representative of the hazards likely to be encountered. The availability of verified and validated models describing the response of energetic materials to various thermal and mechanical threats will enable us to perform parametric studies on systems. This will allow us to estimate their response to hazards that are characteristic of the specific environment experienced by that system, and so to tailor the materials and packaging to minimize risk and maximize performance.