The authors' aim was to examine the relations among mental multiplication, working memory load (WML), and automaticity by alternating the difficulty level of task characteristics. In Experiment 1, involving 30 fifth-grade students with mixed abilities, a 2 (WML) × 2 (automaticity) design was utilized. In Experiment 2, involving 21 high-achieving mathematics learners and 21 low-achieving mathematics learners in Grade 4, a 2 (WML) × 2 (automaticity) × 2 (achievement) design was utilized. Regardless of level of automaticity, individuals under low-WML conditions performed more accurately and faster. Regardless of level of WML, individuals under high automaticity conditions performed more accurately and faster. Group difference was significant. The simple effect of WML was bigger under the conditions with low automaticity, in comparison to the conditions with high automaticity. Alternating difficulty level simultaneously in 2 dimensions of testing conditions posed an amplified impact on the low-achieving group.