This paper is part of a PhD study, its purpose is to differentiate the physics problem solving processes used by between "more successful" and "less successful" problem solvers working on a problem intended to be unfamiliar in type for the problem solvers. The focus of the study was to see if the more successful problem solvers were more metacognitive in their approaches and, if so, what did these differences look like. The problem was given to 21 students, all Physics majors enrolled in a Bachelor of Science degree within the Faculty of Education at Universiti Teknologi Malaysia and data from the most successful 5 and least successful 5 were used for analysis. The chosen instruments for the study were participant observation and interview. Participants were instructed to "think aloud" as they went about solving the problem; they were then immediately interviewed about what they had said and drawn with the goal of further exploring their problem solving strategies. Data was transcribed and then analysed manually using NVivo. From the data, it was found that "more successful" and "less successful" show 4 differences and 1 similarity in solving the mechanics problem. The data shows that more successful engaged in planning more often than less successful, less successful rereads more often than more successful, more successful qualitative analysed the problem more often than less successful, more successful used abstract concept while less successful used more naïve concepts than less successful made naïve concepts to represent variables, monitored their thinking, and evaluating their answer. On the other hand, less successful planning as well as "more successful", leaps into equation, drew the diagram, less spend time making qualitative analysis, often rereads the question, made naïve concepts to represent variables, monitored their thinking, and made no evaluating to their answer.