This research aimed to examine whether the activities related to mental computation strategies had any effect on the development of the second-grade elementary school students' problem-solving (PS) skills. Participants of the study are comprised of 14 children aged 7-8 years and enrolled in the second grade of the elementary school of a village. For the collection of data, the Problem-Solving Task (PST) which is composed of 12 problems linked with addition and subtraction operations was utilized. The identification of problems to be included in PST was based on problem types suggested by Sarama and Clements (2009). In this action research which lasted five weeks and three hours per week, 13 mental addition-subtraction activities (MASAs) were applied to children. Responses given by children to PST were examined in the context of time, accuracy, and strategy variables. At the end of executed activities, it was discerned that mental computation strategies used by children in solving problems got diversified, children's response speed was accelerated and the number of correct answers given by children went up. Before MASAs, it was found that children had difficulty in solving problems through mental computation, gave random answers, or tried to count with their fingers for computation. After MASAs, it was deduced that strategies that were selected by children differed on the basis of problem types.