One or more searchers must capture an invisible evader hiding in the nodes of a graph. We study this graph search problem; we emphasize that we study the capture of a node-located evader, which has received less attention than edge search. We show that in general graphs the problem of node search is easier than that of edge search (however node search is NP-complete, just like edge search). We concentrate on the internal monotone connected (IMC) node search of trees and show that it is essentially equivalent to IMC edge search. For IMC node search on general graphs we present a new algorithm: GSST (Guaranteed Search by Spanning Tree) which clears the graph G by performing all its clearing moves along a spanning tree T of G. We prove the existence of probabilistically complete variants of GSST. Our experiments also indicate that GSST can efficiently node-clear large graphs given only a small running time. An implementation of GSST is also provided and made publicly available.