We had previously developed an improved Ames module to directly determine the mutagenicity of gaseous formaldehyde (HCHO) and toluene without liquid extraction. This study further evaluated the suitability and sensitivity of this module on whole and real polluted air samples. For this, two common brands of stick incense (A and B) and cigarettes (A and B) were harvested, and various types of incense smoke (IS) and sidestream cigarette smoke (SCS) samples were generated by lighting 3, 6, 12, 24, 30, or 36 incense sticks, and by lighting 1, 2, or 3 cigarettes, respectively, in an acrylic box. CO2, CO, total volatile organic compound (TVOC), PM1.0, and HCHO concentrations in the air samples were determined, and all air samples did not partially fit the requirements of the air quality standards. The smoke samples were then directly exposed to TA100 for 10, 20, 30, or 60 min in our exposure module. Exposure to IS (brand A) for 30 to 60 min and exposure to IS (brand B) for 60 min led to statistically (p < 0.05) weak (below the twofold rule) but dose‐dependent mutagenic activities either with or without metabolic activation. Furthermore, a short‐term exposure (10–60 min) to SCS (brands A and B) displayed statistically significant (p < 0.05) direct‐acting, indirect‐acting, time‐ and dose‐dependent mutagenic activities. Furthermore, our data also support that the liver S9 enzyme could enhance the mutagenic activities in most IS and SCS samples. This study confirmed that the modified Ames module can be applied to directly detect the mutagenic activities of real polluted air samples.