Cough-related variables such as cough frequency, time to onset (i.e. the time until the first cough occurs) and the cough index (i.e. the ratio between the cough frequency and the time to onset) may be important when interpreting results of cough challenges for therapeutic interventions or for comparative research purposes. Nevertheless, repeatability (or reproducibility) for these widely used variables has been poorly studied. In thirty normal subjects, coughing was induced by inhalation of threshold (T) and suprathreshold (1.6 x T) concentrations of ultrasonically nebulized distilled water (fog). Cough threshold was taken as the lowest fog concentration that evoked at least one cough effort during two challenges separated by a 30-min interval. During challenges performed at both threshold and suprathreshold stimulus intensity, cough frequency, time to onset, and the cough index were assessed; within-subject repeatability for these variables was subsequently evaluated. Median +/- interquartile range cough threshold value was 0.9+/-0.5 mL x min(-1). During the two challenges performed to assess cough threshold, the mean +/-SD values of cough frequency, time to onset, and cough index were similar (5.0+/-2.7 and 5.3+/-3.1 coughs x min(-1), 32.4+/-13.3 and 32.9+/-13.6 s, and 0.2+/-0.2 and 0.2+/-0.2, respectively). However, none of these cough-related variables proved to be sufficiently repeatable. During the two challenges performed at suprathreshold stimulus intensity, mean values of cough frequency, time to onset, and cough index were also similar (20.0+/-9.0 and 18.2+/-10.2 coughs x min(-1), 13.5+/-5.8 and 12.0+/-4.62 s, and 1.7+/-1.0 and 1.8+/-1.2); furthermore, all considered variable of suprathreshold challenge turned out to be reproducible. In conclusion, during fog challenges at threshold stimulus intensity, cough frequency, time to onset and cough index cannot reliably be used for evaluating cough responses. However, these cough-related variables may represent useful and reliable research tools in the evaluation of suprathreshold cough responses.