Testing and certification of packagings for transport of dangerous goods demand quite a lot of resources, economical as well as personal. The main principle is that every type of packaging to be used for dangerous goods should be tested. Aiming to minimize this amount on testing two possibilities for selective testing have been studied in this report: when the packaging differs from a type earlier tested only because some geometric dimension is smaller, or when it comes to combination packagings. In the report it will be shown that selective testing is possible in the case of dimension variations when the reduction is less than around 10%. In the report it is furthermore demonstrated that selective testing is possible for combination packagings. For combination packagings it is proposed that the inner packaging be classified in a class depending on its damage boundary (as decided by drop test). The combination is then tested with one type of inner packagings (classified). If the tested packaging passes the test, it is then approved for all packagings in that class and higher classes without further tests. To verify this possibility a mathematical model is developed for the dynamic load between the packagings. By instrumented drop tests it is then demonstrated that the dynamical influence is relatively small.