This paper reports the results of an experimental rating task testing the acceptability of variation in Modern Hebrew spirantization. Modern Hebrew spirantization is manifested by the alternation of the stops [p], [b], and [k] with their fricative counterparts [f], [v], and [x], in which fricatives occur post-vocalically and root-finally, and stops occur elsewhere, as in (1). (1) /p/ to [f] /pgS/ [lifgoS] [pagaS] ‘to meet’ /b/ to [v] /bgd/ [livgod] [bagad] ‘to betray’ /k/ tp [x] /ktb/ [lixtov] [katav] ‘to write’ Due to historical sound mergers and more recent borrowings, there are acoustically identical stops and fricatives that do not alternate and which may surface as exceptions with respect to their environment (i.e. stops in post-vocalic position, fricatives in word-initial position). In addition to these cases of exceptional non-alternation, variation in alternating sounds has also been reported (Adam 2002, Temkin Martínez 2008). As with the non-alternating segments, variation manifests itself as stops and fricatives surfacing in positions where they are not expected given spirantization. It has been argued that this variation stems from the high frequency of non-alternating segments, driving alternating segments towards non-alternation (Adam 2002). To examine the nature of variation in Modern Hebrew spirantization, an experimental rating task was conducted with 74 native speakers. The results show that variation is acceptable not only in alternating segments, but in non-alternating segments as well. However, variation is significantly more acceptable in alternating segments than in non-alternating ones, suggesting that speakers still distinguish between alternating and non-alternating segments. Additionally, trends in participants’ acceptability of variation reveal that it is impossible to determine directionality as to whether alternating segments are becoming non-alternating.