The use of air as the caloric medium was investigated and compared to that of water. The ability of air and water to generate surface temperatures close to the fluid outlet temperature was studied by irrigation onto a thermocouple. Results of caloric tests performed with air and water are compared by calculating the average nystagmus response for the hot and cold stimuli for two groups of sixty patients. Also, the number of canal paresis, directional preponderance, mixed, bi-lateral canal paresis and normal responses are compared for each group. Finally, the effect of performing a bi-thermal air caloric on a patient having a unilateral drum perforation is discussed.