Arithmetic tie effect is a phenomenon that tie problems (e.g. 3+3 or 6×6) are solved faster than non-tie problems (e.g. 3+4 or 6×7). The purpose of this study was to compare the predictions of three hypotheses on the tie effect: learning frequency, encoding advantage, and two-factor model. The utilization of the true-false evaluation, from which the participants ( n =22) were required to ascertain whether an equation (e.g. 3+4=7) was true, revealed that the tie effect was larger in addition than in multiplication. This result supported the two-factor model and led to further discussion on the process of arithmetic verification.