This study aims to compare the effects of activity-based instruction and traditional instruction on fifth grade primary school students' conceptual development of probability. The study was conducted through quasi-experimental method and carried out with 50 5th grade primary school students, 25 for experimental group and 25 for control group. The Conceptual Development Test consisting of 12 open-ended questions was administered to the students in the study before and after the treatment. Data were analyzed using independent samples t-test, and analysis of covariance (ANCOVA). It was determined that activity-based instruction was more effective in helping students develop the probability concepts than did traditional one.