research should be restricted to P3 and P4 facilities. If NIH 
believes that any recombinant organism is too hazardous to be 
released into the environment intentionally, then it should not 
permit research in facilities where accidental release is 
inevitable . 
C . The System of Classifying Experiments On The 
Basis of Risk is Too Elaborate to be Workable 
Under the NIH Guidelines, a proposed recombinant DNA 
experiment must be assigned to one of thirteen different contain- 
ment levels depending on the hazard it poses to humans. It must 
be assigned one of four physical containment levels and one of 
three biological containment levels or it may be prohibited. 
We do not know enough about the hazards of recombinant DNA to 
make these finely tuned evaluations of risk. Since recombinant 
DNA molecules are placed in living host cells, which have the 
potential of surviving and multiplying in the environment, even 
one experiment, which is mistakenly considered harmless, could 
cause widespread, irreversible damage. Therefore, the classifica- 
tion system should be designed to protect us against our ignorance 
and recognize that our estimates of risk of different experiments 
are very crude. 
The draft EIS repeatedly states that we cannot predict 
the effects of any particular genetic recombination. 
21 
Appendix K — 147 
