One important aspect of both shadowing and simultaneous interpreting is the ear-voice-span (EVS). This refers to the lag elapsing between the subject’s or interpreter’s monitoring of the input, and his actually repeating or encoding it. The EVS is, as one may expect, longer in simultaneous translation than in shadowing. In this paper, the author has concentrated on the EVS aspect of simultaneous translation. Some of the main questions to be answered with respect to EVS are: 1) what is the length and nature of the segments the interpreter needs to monitor before he can start encoding? And 2) how does he segment the input? This paper is divided into two parts to examine these questions separately.