We have used stable transformation of a cloned Drosophila hsp70 gene into mouse Ltk- cells as an assay system to determine which sequences are involved in the temperature-induced transcription of the gene. We have analyzed the effect of different external deletions on the ability of the gene to be transcribed after temperature elevation. We have also constructed chimeric genes containing different sequences of the hsp70 promoter region joined to the herpes simplex virus thymidine kinase gene, and we have studied the heat-shock inducibility of the thymidine kinase gene after transformation into mouse cells. Our results suggest that the heat-shock genes are under negative control and that their activation is the result of a derepression of the gene. In addition, our experiments indicate that the heat-shock consensus sequence is necessary but not sufficient for temperature-induced transcription, and its role seems to be functionally similar to other transcription signals (i.e. the CAT box) located upstream from the TATA box in several eukaryotic genes. We are proposing that there are two additional regulatory elements besides the consensus sequence which must be present for proper temperature-dependent transcription to occur. Either one of these elements, which are located on opposite sides of the TATA box, may serve at any one time to ensure transcription; that is, only one at a time is needed for proper expression of the heat-shock gene.