Since the question asks for the time complexity we should consider both the average and the worst time complexities. In a binary search tree the values are sorted such that for each node the elements bigger than it are placed in the right side and elements smaller than it are placed in the left side. This means that the searching algorithm performed when a certain value in the tree is searched for compares the value to each node to see whether the left or right subtrees should be searched in. This significantly reduces the time complexity of the search operation, averaging to O(log n).
Worst case scenario, a binary tree has the maximum (or minimum element) at the root and all the other elements are sorted on one side of the root, effectively turning the tree into a linked list. In this particular case the search operation in a binary search tree will have the time complexity of O(n) because it will iterate through all the elements like in a regular linked list.