OBJECTIVE Toward understanding the function of sleep spindle, we examined whether sensory stimulation triggers sleep spindles. METHODS Eleven normal subjects participated in the experiments. METHODS The subjects had a nap in the afternoon, and sensory stimulation was applied during sleep stage 2. METHODS 21-channel EEG was recorded during the 2-3 hour experiment carried out between 13:00 and 16:00. Somatosensory, auditory, or visual stimulation was performed over a 5-minute period during stage 2. The frequency and duration of spindles were compared in 2 different segments of 5 minutes, with and without sensory stimulation. The latency from the onset of a sensory stimulus to the succeeding spindle was also analyzed. To estimate the active brain regions during a spindle, the EEG recordings were modeled with a single equivalent moving dipole (SEMD) model. RESULTS In the period with stimulation, spindle frequency and duration increased compared with the period without stimulation. Statistical tests revealed that with stimulation, the interval between 2 consecutive spindles was significantly shorter (p < 0.05, regardless of the modality) and that the duration of the spindles was significantly longer with stimulation (p < 0.05, regardless of the modality). The latency was approximately 2 s. During a spindle after somatosensory stimulation brain activities were observed near the somatosensory area, while with auditory stimulation active regions were observed near the auditory cortex. CONCLUSIONS A sensory stimulus appeared to trigger a sleep spindle during sleep stage 2. SEMD trajectories suggest that active brain regions during spindle are different according to the modality of the preceding stimulus.