In T 315/03 (OJ 2006, 15) the board held that, in animal manipulation cases, the test in T 19/90 (OJ 1990, 476) was appropriate. This differed in several respects from the test in R. 23d(d) EPC 1973 (R. 28(d) EPC), most importantly by allowing matters other than animal suffering and medical benefit to be taken into account. Whereas the R. 23d(d) EPC 1973 test only required a likelihood of animal suffering, however minor, and a likelihood of achieving a medical benefit, the test in T 19/90 required a "careful weighing up" of the matters to be balanced. This clearly allowed an appraisal of animal suffering, of environmental risks and of the feasibility of using non-animal alternatives.

A wider range of benefits to mankind than the substantial medical benefit found in R. 23d(d) EPC 1973 could also be considered. Since the test in T 19/90 was "mainly" the basis of assessment, further arguments as to the appropriate standard of morality or "ordre public" could additionally be considered, but all arguments had to be supported by evidence.

In T 315/03 the board found that claims directed to rodents failed the R. 23d(d) EPC 1973 test and had therefore to be refused, as the patent disclosed a likelihood of animal suffering but not a likelihood of medical benefit when applying the invention to this generic class of animals. The same conclusion would have been reached under an Art. 53(a) EPC 1973 assessment. However, both tests were satisfied when the invention was restricted to mice.