An enzyme-linked immunosorbent assay (ELISA) for measurement of levels of heat shock proteins 73 and 72 (HSP73/72) in cultured cells and tissues is described. The assay involves detection of HSP73/72 in cell homogenates in 96-well plates using a specific monoclonal antibody. The assay has been used to explore the relationship between the amount of HSP73/72 in a cell and its response to heat shock, both before and after the development of thermotolerance. Six mammalian cell lines with differing responses to heat were characterized with respect to their response to heat treatments at 44 degrees C and concentrations of HSP73/72. Contrary to the widely expressed idea that the amount of HSP73/72 dictates the degree of heat resistance, no positive correlation between levels of HSP73/72 and heat resistance was found for the six lines tested here: if one particular line, a mutant selected for heat resistance, was excluded from the analysis, there was a negative correlation between HSP73/72 levels and heat resistance. A different result was, however, obtained when thermotolerant (transiently resistant) cells were compared to control cells. Here, we found a good correlation between the extent of thermotolerance and the amount of HSP73/72, suggesting that an increase in HSP73/72 level is important for the development of thermotolerance. The validity of the ELISA technique was checked using a second method for quantifying levels of HSP73/72. This involved uniform radiolabeling of cellular proteins, separation on two-dimensional gels and radioscanning to quantify radioactivity in each protein. The second technique is more powerful in that different isoforms of HSP73/72 can be distinguished, but it is more difficult to perform, is more labour intensive and requires an expensive device for gel scanning. The results using the second technique agreed well with those from the immunoassay and indicated that the level of the highly inducible HSP72 correlated best with the extent of thermotolerance.