The aim of this experiment was to determine whether ventilatory measurements in adult restrained mice provide a valid assessment of chemosensitivity. We used whole-body plethysmography to compare breathing patterns in eight restrained and eight unrestrained outbred Swiss mice during air breathing, hypercapnia, and hypoxia. The mice in the restrained group were each placed in a loosely restraining wire-mesh cage. The unrestrained mice could move freely inside the plethysmograph. All the mice received three hypercapnic stimuli (8.5% CO2) and three hypoxic isocapnic stimuli (10% O2, 3.5% CO2). As compared to unrestrained mice, restrained mice had significantly lower breath durations (TT, 445+/-110 ms vs. 323+/-32 ms) and higher ventilation (VE) levels (15.7+/-2.6 microl/(sec x g) vs. 22.2+/-4.5 microl/(sec x g)), whereas no difference was observed for tidal volume (VT). The increases in frequency and ventilation from baseline to hypercapnia were not significantly different in restrained and unrestrained mice. The VE response to hypoxia was marginally higher in restrained mice. We conclude that chemosensitivity to hypercapnia, and to a lesser extent to hypoxia, can be measured in restrained adult mice, but that the baseline breathing pattern cannot.