We expanded a previously described rule-based computerized method to recognize the sensory irritating effect of airborne chemicals. Using 2-chlorobenzylchloride (CBC) as a sensory irritant, characteristic modifications of the normal breathing pattern of exposed mice were evaluated by measuring the duration of braking (TB) after inspiration and the resulting decrease in breathing frequency. From the measurement of TB, each breath was then classified as normal (N) or sensory irritation (S). Using increasing exposure concentrations, the classification S increased from < or = 2% (equivalent to sham-exposure) to 100% within a narrow exposure concentration range. The potency of CBC was then evaluated by calculating the concentration necessary to produce 50% of the breaths classified as S, i.e., S50. This approach is easier to use than obtaining RD50 (decrease in respiratory frequency by 50%) when high exposure concentrations are difficult to achieve. Detection limits were also established for this bioassay and experiments were conducted to obtain a level of response just around these limits, in order to delineate the practicality of using this bioassay at low exposure concentrations. Using this approach, sensory irritation was the only effect induced by CBC at low exposure concentrations. However, bronchoconstriction and pulmonary irritation were superimposed on this effect at higher exposure concentrations.