A ship, advancing in a level ice field with continuous icebreaking mode, is subject to several components of ice resistance such as resistance due to breaking of ice, ice buoyancy, removing ice and ship motion. Although the third and forth components increase with ship speed, the first component, i. e. icebreaking resistance, is the most significant. This paper studies the icebreaking resistance at low ship speed. The results of model experiments on icebreaking resistance were analyzed in detail for three different bows. They were a conventional wedge-shaped bow, spoon bow and concave bow. The following results were obtained. (1) Independently of the bow shape, icebreaking resistance is in proportion to the total crack length, ice flexural strength and ice thickness squared. (2) Decreasing hull stem angle at ice contact point raises the width of broken ice piece, resultingin shorter crack length. The icebreaking resistance nondimensionalized by the ice flexural strength and ice thickness is determined by the tangent of hull stem angle at the contact point. (3) Since the spoon bow has smaller stem angle and larger waterline angle, the total crack length becomes shorter compared to the other bows. This is the reason why the spoon bow gives the lowest ice resistance.