Sindbis virus (SIN) contains an in-frame opal termination codon in the nonstructural protein-coding region separating nsP3 and nsP4 and provides a useful tool to study the readthrough phenomenon of the termination codon in host cells and its role in viral replication. We have changed the opal codon by site-directed mutagenesis of a full-length SIN cDNA clone to either sense amino acids (serine, tryptophan, or arginine) or the other two translation termination codons (amber or ochre). Transcripts from all of the mutant cDNA clones were infectious when used to transfect chicken embryo fibroblasts. The resulting progeny virus stocks were then used to study the effects of these mutations on viral protein and RNA synthesis, growth properties, host range, and fitness compared with the parental strain. None of the mutants showed temperature sensitivity in plaquing efficiency or plaque morphology on chicken embryo fibroblast monolayers. Relative to the wild-type parent, the mutants containing sense replacements overproduced nsP34 but not nsP4 and made slightly decreased levels of nsP3, with a delay in its appearance. This indicates that the cleavage separating nsP3 and nsP4 occurs in these mutants and also that the level of nsP4 is not regulated solely by readthrough of the opal codon. The amber and ochre mutants produced decreased levels of nsP34, and the ochre mutant grew significantly more slowly than the other mutants or wild-type virus. For all five mutants, RNA synthesis early in infection was inhibited compared with that of the parental virus. This effect was apparent at multiplicities of infection of 20 PFU per cell but not at 100 PFU per cell. Using in situ hybridization to distinguish between mutant and wild-type plaques, we have studied the behavior of the serine mutant in a high-multiplicity growth competition experiment with wild-type virus. The wild-type virus eventually outcompeted the mutant after several passages, and these results indicate that this mutation has resulted in effects that are at least partially cis acting. Furthermore, by studying the growth, plaque formation, and protein synthesis of the mutants in various cell types, we have observed host range effects of the mutations, especially in mosquito and human cells. In addition, we have demonstrated, at least indirectly, that opal, amber, and ochre termination codons in the SIN nucleotide context can be suppressed in cultured cells of chicken, human, hamster, and mosquito origin.