Abstract.-Two new nonsense suppressors in Escherichia coli were found in partial diploids carrying F'14 and were shown to be on the episome. These suppressors can exist only in cells which also contain the su-allele, i.e., su+/suheterozygotes. Presumably the mutations cause an alteration of an essential cellular component, the complete loss of which is lethal. Su7, an amber suppressor, has an efficiency of 76 per cent and su8, an ochre suppressor, an efficiency of 4 per cent.Extensive searches by several workers have revealed only three amber suppressors in Escherichia coli.'-3 Such suppressors are thought to result from changes in the anticodon of a transfer RNA (tRNA) allowing the altered tRNA to pair with UAG, the amber triplet, at the ribosome.4-6 Seven amino acids have codons related to UAG by a single base change: therefore, it should be possible to generate at least one suppressor for each of these amino acids by a mutation affecting the anticodon. The fact that no suppressors have been found which insert some of these amino acids (tryptophan, lysine, and glutamic acid) suggests a fundamental restriction on the generation of suppressors.7 Assuming that anticodon changes in certain tRNA's do not prevent aminoacylation, the most likely restriction is that the tRNA which might become the suppressor is indispensable, and its loss is lethal to the cell. If this were true, such potentially lethal mutations affecting tRNA's should be recoverable in cells which are diploid for these genes. Such strains can retain one copy of a duplicated indispensable tRNA gene to perform its normal function, allowing the second to mutate to yield a suppressor. Other potential mutations leading to suppression but not involving tRNA's might also be lethal to the haploid cell but could be revealed in partial diploids. This report presents a general approach for the isolation of such recessive lethal suppressors and describes a new amber and a new ochre suppressor of this type.Materials and Methods.