Ongoing studies are identifying and quantifying the volatile reaction products of gas-phase compounds present in the indoor environment, especially dicarbonyls, and investigating immunotoxic and hypersensitivity effects of these reaction products in both in vitro and in vivo models. Utilizing the Vitrocell? exposure module, human lung epithelial (A549) cells were exposed to clean air, air contaminated with limonene, or a mixture of limonene and ozone. Modifications of the growth and culturing technique for the A549 cells were made to optimize the performance of the cell line. Exposure to limonene resulted in significant increases in IL-8 and MCP-1 at 24 hours post exposure when compared to the clean air control. Exposure to limonene + ozone resulted in a decreased production of MCP-1 at the 10 and 24 hour time points and viability when compared to limonene alone. The effect of pre stimulation on cell viability and pro inflammatory cytokine production was also investigated following exposure to limonene, or limonene + ozone. In contrast to unstimulated cells, no changes in cell viability or cytokine production were observed. Lower exposure concentrations of limonene and ozone were also used to explore the influence of exposure concentration of the pro inflammatory and viability responses of A549 cells. Pulmonary epithelial cells were exposed to clean air, limonene (500 ppb), or limonene (500 ppb) + ozone (100 ppb). In contrast to previous studies that used high concentrations of limonene (20 ppm) and ozone (5 ppm), no changes in cell viability or cytokine production were observed following exposure. This result emphasizes the importance of chemical exposure concentration and raises the issue of hazard identification. Since A549 cells can only be exposed on a single occasion in the current system the relevance of using a single exposure of realistic chemical concentrations verses a single exposure of a high concentration of chemical is being explored. These results suggest the importance of exposure time, chemical concentration, pre stimulation, time point for collection of endpoint, and endpoint sensitivity. These findings have identified potential sources of variability for these types of analyses and emphasize the need for standardization.