In this paper we prove a Markov Theorem for the virtual braid group and for some analogs of this structure. The virtual braid group is the natural companion to the category of virtual knots, just as the Artin braid group is to classical knots and links. In classical knot theory the braid group gives a fundamental algebraic structure associated with knots. The Alexander Theorem tells us that every knot or link can be isotoped to braid form. The capstone of this relationship is the Markov Theorem, giving necessary and sufficient conditions for two braids to close to the same link (where sameness of two links means that they are ambient isotopic).The Markov Theorem in classical knot theory is not easy to prove. The Theorem was originally stated by A.A. Markov with three moves and then N. Weinberg reduced them to the known two moves [29,38]. The first complete proof is due to J. Birman [4]. Other published proofs are due to D. Bennequin [3], H. Morton [30], P. Traczyk [35] and S. Lambropoulou [24,25]. In this paper we shall follow the "L-Move" methods of Lambropoulou. In the L-move approach to the Markov theorem, one gives a very simple uniform move that can be applied anywhere in a braid to produce a braid with the same closure. This move, the L-move, consists in cutting a strand of the braid and taking the top of the cut to the bottom of the braid (entirely above or entirely below the braid) and taking the bottom of the cut to the top of the braid (uniformly above or below in correspondence with the choice for the other end of the cut). See Figure 15 for an illustration of a classical L-move. One then proves that two braids have the same closure if and only if they are related by a sequence of L-moves. Once this L-Move Markov Theorem is established, one can reformulate the result in various ways, including the more algebraic classical Markov Theorem that uses conjugation and stabilization moves to relate braids with equivalent closures.Up to now [25,10,26] the L-moves were only used for proving analogues of the Markov theorem for classical knots and links in 3-manifolds (with or without boundary). Our approach to a Markov Theorem for virtual knots and links follows a similar strategy to the classical case, but necessarily must take into account properties of virtual knots and links that diverge from the classical case. In particlular, we use L-moves that are purely virtual, as well as considering the effect of allowed and forbidden moves of the virtual