Sensory funneling of liminal multiple-point air-puff stimulation on the skin was evaluated by measuring somatosensory evoked potentials (SEPs), reaction time (RT) and subjective detection threshold. Single-point air-puffs at the threshold intensity, and 2-point and 3-point air-puffs (10 mm apart) of the same intensity were delivered to the volar hand for recording SEPs, RT and detection threshold. Two- and 3-point stimulation produced a funneled sensation which feels stronger than the sensation produced by the single-point stimulation. Thus, the detection threshold was lower and the detection probability high with multiple-point stimulation. The latency of P300 SEP component showed a trend for shortening, while the P300 amplitude was reduced with multiple-point stimulation. However, these differences did not reach statistical significance. In sharp contrast with the relatively invariant SEP measures, the RT showed a dramatic reduction with multiple-point stimulation. The results suggest that P300 SEP component is less affected by the funneling of sensory input but is related to the detection of the threshold stimulus in an all-or-none fashion. In contrast, the RT is linearly related to the funneling of input and the resultant higher detection probability. We speculate that an intermediate neural process between the P300 elicitation and the motor command was facilitated by the funneling of multiple air-puffs.