Twenty-one commercially available chemical steam sterilization indicators were processed in an empty autoclave for various times at temperatures between 240 and 270 degrees F (ca. 116 and 132 degrees C). The time required to reach a sterilized reading at each temperature was plotted on a semilogarithmic time-temperature plot and compared with the time-temperature sterilization curve for Bacillus stearothermophilus. Five of the indicators had time-temperature kinetics similar to those of B. stearothermophilus, but three of these overestimated the effect of processing. Two of the indicators overestimated the effect of processing and were less sensitive to temperature changes when was B. stearothermophilus. Thirteen of the indicators had time-temperature curves that crossed the B. stearothermophilus plot. One indicator produced such ambiguous results that no determinations could be made with it. Out of 21 indicators tested, only 2 appear to be capable of accurately integrating the time-temperature effect at temperatures between 240 and 270 degrees F. The other indicators should be used only after careful analysis of their suitability for use at a given temperature.