Summary Temperature sensitive products are often transported in non-refrigerated systems, protected from environmental temperatures by thermal insulation and the provision of a source of cold inside the package. This work presents a method for optimising the design of thermally insulated boxes for lowest cost (while ensuring that product temperatures are maintained within set limits) by combining a heat transfer model with genetic algorithm optimisation. An example optimisation problem is presented in which a temperature sensitive product is transported across the US and must be maintained between −1 and 8 °C. Independent optimisation operations successfully identified solutions that were very similar, though not identical, thereby providing confidence in the approach to determine optimal solutions to complex problems. The system developed provides a rapid method to optimise the design of an insulated box for minimum cost while maintaining the product in the appropriate temperature range.