This study examined if and how thinking time, defined as the time from the start of the problem to the students' first action to solve, influences their problem-solving efficiency, defined as the number of steps taken to solve the problem, using Graspable Math, a web-based mathematical tool. We found that: (1) students who took longer to think about the problems solved individual problems with fewer steps; (2) thinking time significantly predicted step efficiency above and beyond students' algebra knowledge or their affective disposition towards mathematics; and (3) the relation between thinking time and step efficiency was not moderated by students' knowledge in algebra, math self-efficacy, or math anxiety.