We have attempted to define a just detectable effect (JDE) for three different types of reactions along the respiratory tract: (a) sensory irritation of the upper airways (S), (b) airflow limitation along the conducting airways (A), and (c) pulmonary irritation at the alveolar level (P1 or P). Each type of reaction, S, A, P1 or P, was recognized by analyzing the breathing pattern of unanesthetized mice held in body plethysmographs. A rule-based computer program analyzed each breath during a period of 3.75 h and classified each breath as normal (N) or falling in any of the above categories (i.e., S, A, P1 or P). Eight groups of four mice were used for sham exposures: exposed to water vapor. These data sets were used, as sham exposure data, to define the variation which can occur with time in order to define an expected range of normal variation. Once this range was established, we defined JDE values for each type of effect and used such values to evaluate the results obtained in exposed animals. Eight groups of four mice were exposed to a mixture of airborne chemicals, machining fluid G (MFG), at concentrations from 0.17 to 55 mg/m3. Data sets for individual animals and for each group of animals exposed to MFG were analyzed to determine if and when a particular effect occurred. It was possible to recognize the effects of low exposure concentrations on groups of exposed animals or individual animals within each group. This procedure will be valuable when investigating the effect of airborne chemicals and when it is impossible to generate high exposure concentrations to define concentration-response relationships.