We study the multicut on trees and the generalized multiway Cut on trees problems. For the multicut on trees problem, we present a parameterized algorithm that runs in time O * (ρ k ), where ρ = √ 2 + 1 ≈ 1.555 is the positive root of the polynomial x 4 −2x 2 −1. This improves the current-best algorithm of Chen et al. that runs in time O * (1.619 k ). For the generalized multiway cut on trees problem, we show that this problem is solvable in polynomial time if the number of terminal sets is fixed; this answers an open question posed in a recent paper by Liu and Zhang. By reducing the generalized multiway cut on trees problem to the multicut on trees problem, our results give a parameterized algorithm that solves the generalized multiway cut on trees problem in time O * (ρ k ), where ρ = √ 2 + 1 ≈ 1.555 time.For convenience, we will refer to a pair of terminals (u i , v i ) ∈ R by a request, and we will also say that u i has a request to v i , and vice versa. generalized multiway cut on trees (GMWCT) Given: A tree T and and a collection of vertex/terminal-sets S 1 , . . . S r Parameter: k Question: Is there a set of at most k edges in T whose removal disconnects each pair of vertices in the same terminal set S i , for i = 1, . . . , r?As the name indicates, the GMWCT problem generalizes the well-known multiway cut on trees problem in which there is only one set of terminals.The MCT problem has applications in networking [7]. The problem is known to be NPcomplete, and its optimization version is APX-complete and has an approximation ratio of 2 [11]. Assuming the Unique Games Conjecture, the MCT problem cannot be approximated to within 2 − ǫ [13]. From the parameterized complexity perspective, Guo and Niedermeier [12] showed that the MCT problem is fixed-parameter tractable by giving an O * (2 k ) time algorithm for the problem. (The asymptotic notation O * (f (k)) suppresses any polynomial factor in the input length.) They also showed that MCT has an exponential-size kernel. Bousquet, Daligault, Thomassé, and Yeo, improved the upper bound on the kernel size for MCT to O(k 6 ) [2], which was subsequently improved very recently by Chen et al. [3] to O(k 3 ). Chen et al.[3] also gave a parameterized algorithm for the problem running in time O * (1.619 k ).The multiway cut on trees problem (i.e., there is one set of terminals) was proved to be solvable in polynomial time in [6,8]. Chopra and Rao [6] first gave a polynomial-time greedy algorithm for the problem. More recently, Costa and Billionnet [8] proved that multiway cut on trees can be solved in linear time by dynamic programming. Very recently, Liu and Zhang [15] generalized the multiway cut on trees problem from one set of terminals to allowing multiple terminal sets, which results in the GMWCT defined above. They showed that the GMWCT problem is fixed-parameter tractable by reducing it to the MCT problem [15]. Clearly, the GMWCT problem is NP-complete when the number of terminal sets is part of the input by a trivial reduction from the MCT problem. Liu and Zhan...