Abstract: It is well known that human beings do not have the organs to feel time, but they often use the time information to adapt to the environment in daily life, which may mean that human beings have a certain time processing mechanism. Human beings, in life, can be a wide range of time, from tens of milliseconds to several decades. In the context of this time, the range of the duration cognitive mechanisms in milliseconds that are closely related to the life of the people is one of the focus of the researchers, such as playing basketball, playing music, etc.. Since the beginning of the eighty or ninety's in twentieth Century, researchers have gradually shifted the interest in the study of "duration" to "segmentation", and the emergence of two kinds of hypotheses: the non - segmental hypothesis and the hypothesis of segmentation. In the former, it is argued that there is no segmentation of time cognition, and the scalar timing model can be used to explain the different duration mechanism (Church, 1984). Second, time cognition exists segmentation, and the critical segmentation point may be involved in the 2~3s, 1setc. (Lewis, 2003; P?ppel, 1997). In recent years, there are many researches in the difference between the processing mechanism in 1 second and that more than 1 second (Cellini, Fabbri, Rammsayer, 2014.,et.al), but there is little researches in the duration cognition segmentation in 1 second ( Rammsayer & Lima, Rammsayer, 1991; &Ulrich, 2005). Based on this, this study continues to explore the segmentation of time in one second, which, the important theoretical significance is, not only helps to reveal the processing mechanism of time, but also provides important practical guidance for the practical activities of human, the practical significance is, using time information to adapt to the objective world. The time duration cognition segmentation means that the difference between the different processing mechanism and the representation is different. In the past, there were still differences in characteristic of the time cognition segmentation in 1 second in the past researches, so the study designed the experiment 1 and experiment 2. And in Experiment 1, dual task paradigm was used to investigate the selective interference between 100ms and 1000ms listening interval in mental arithmetic task. The results show that, mental arithmetic task affect the difference threshold of 1000ms listening interval, and the larger mental load is, the greater difference threshold is, but there is no influence on the listening interval difference threshold in 100ms. To verify the results of Experiment 1, experiment 2 used dual task paradigm to investigate the selective interference between 100ms and 1000ms viewing interval in mental arithmetic task. The results show that, mental arithmetic task affect the difference threshold of 1000ms viewing interval, and the larger mental load is, the greater difference threshold is, but there is no influence on the viewing interval difference threshold in 100ms. Two experiments show that the audio and visual interval cognition exist segmentation in 1 second.