Three main issues are touched upon in this paper: (a) main morphological types of Georgian verbs in correlation with verb form, on one hand, and with the categories of telicity, stativeness, and transitivity, on the other; (b) peculiarities of case alternating objects of medial verbs; (c) and formal features of medial verbs as the basis for identification of the medial verb class (along with semantic features). This paper is based on Hopper & Thompson’s (1980) hypothesis that considers transitivity not as a strict dichotomy but as a continuum where various degrees of semantic transitivity may be distinguished. Differences expressed on semantic and syntactical levels mark one class of Georgian verbs, Class III, as intermediate between transitives and intransitives. Labile transitives, a term suggested by Melikishvili (2001), is used in this paper to indicate peculiarities of Class III verbs with respect to transitivity. This term expresses the nature of Class III verbs in terms of transitivity more accurately than the labeling of these verbs as intransitive, transitive, stative voiceless, or active intransitive verbs. It is demonstrated in this paper that various degrees of semantic transitivity as well as the category of telicity have their morphological expression in Georgian. Semantically, a case-alternating object of labile transitive verbs is often an inconcrete, non-definite object and/or is not affected by the action described in the verb. Syntactically, the difference is expressed by the lack of ability to assign a third person object marker to a verb. Case alternating objects of both Class I and Class III verbs are marked by the dative case in the tenses of series I. But the difference is that the dative marked, case alternating object of telic verbs (Class I) is usually marked by a prefix on verbs in old Georgian and in several modern dialects of East Georgia (ს-თლის s-tlis ‘s/he is peeling it’). In contrast, the dative marked, case alternating third person object of Class III verbs (atelic, labile transitives) cannot trigger a person marker on a verb (თამაშობს tamashobs but not *ს-თამაშობს *s-tamashobs ‘s/he is playing’). The latter forms are not attested, not even in those dialects where the person marking of the dative marked third person direct object is still preserved. As is described above, besides semantic peculiarities, this verb class, standing in an intermediate position between transitives and intransitives, exhibits morphological and syntactical features of its own.