Attempts to develop baculovirus-based insecticides by insertion of genes encoding enzyme inhibitors, neuropeptides or toxins have met with some success. However, it is often difficult to ensure correct processing or secretion of the encoded peptides. Here we tested a simpler strategy by insertion of an antisense fragment of a host gene to block translation of a protein essential for larval growth and development. We selected the c-myc gene for two main reasons: (i) its protein is known to be well conserved in evolution and to have multiple essential functions during development; and (ii) c-myc family genes have yet to be characterized in insects, thus blockage of essential genes by anti-sense transcripts from a strong virus promoter could provide a sensitive test for the existence of myc-like gene products. An appropriate fragment of the human c-myc gene was inserted downstream from the polyhedrin promoter of Autographa californica nucleopolyhedrovirus and tested in bioassays on Spodoptera frugiperda larvae. Western blot analysis with a human c-myc antibody revealed an endogenous protein band which bound specifically to these antibodies. This band disappeared more rapidly from cells infected with the antisense c-myc recombinant virus than from those infected with c-myc-negative virus. Results of bioassays showed that the antisense construct stopped feeding as soon as the polyhedrin promoter-driven transcripts accumulated, followed shortly by death of the larvae. These results suggest that c-myc-like protein(s) exist in insects and that the antisense strategy is an effective approach to virus insecticide productions.