The aim of this study was to investigate the effects of activity based learning on sixth grade students' mathematics achievement in comparison to traditional learning, and determine their attitudes towards activities. Experimental research design with pretestposttest control group was applied in the study. The participants, which included 78 sixth grade students, were previously divided into two groups by random assignment. One group received traditional learning and the other group received activity based learning. The study was carried out for four weeks on the mathematics subject of "integers". For collecting data the mathematics success test and the Likert type attitudes towards mathematics activities scale were used as pretest and posttest. As a result, the academic achievement for both groups were positive increased. On the other hand, while the attitudes towards activities of students' in the experiment group decreased significantly, they increased in the control group.