Studies of vestibular recruitment using the caloric test have not been tried in general because repetition of irrigation is time-consuming and results are influenced by adaptation. However, we thought it would be possible to detect vestibular recruitment by evaluating how caloric responses increased from onset to peak or decreased from peak to end. Slow-phase velocities of nystagmus were transformed to logarithms to evaluate the coefficient of the upslope or downslope line. If the absolute value of the coef-ficient of the slope was larger, the response was considered to have reached the maximum quickly and to have shown the recruitment phenomenon. In an upslope after 44 o irrigation, the average coefficient of ipsilateral responses in patients with vestibular disorder (3.10+/-1.36x10 - 2 ) was significantly larger than that of responses in normal subjects (1.68+/-0.45x10 - 2 ) and relatively larger than that of contralateral responses in patients with vestibular disorder (2.03+/-0.70x10 - 2 ), and of ipsilateral responses in patients with retrolabyrinthine disorder (-2.03+/-1.14x10 - 2 ). In an upslope after 30 o irrigation and in a downslope after irrigation at either temperature, no tendency was noted. Vestibular recruitment seemed to be measurable in the upslope after warm irrigation.