This research has aimed to study the effectiveness of training in cognitive-metacognitive strategies upon the cognitive load and working memory of senior primary school students with specific learning difficulties in reading by means of a semi-experimental method and using pre-tests and post-tests with a control group. The statistical population consisted of senior primary school students from the city of Miandoab in the school year of 2018-2019; using a random multi-stage sampling method, 30 students with specific learning difficulties in reading were selected and randomly divided into test and control groups. To collect data, The Wechsler Adult Intelligence Scale (Fourth Edition), whose content validity has been verified by experts and shown to have a split-half reliability coefficient of 0.71 to 0.86, the Dyslexia Symptoms Checklist devised by Mikaili and Farahani (2005), whose validity has been verified by experts and has a reliability of 0.92, the Paas Cognitive Load Test (1994), whose validity has been confirmed by experts and has a reliability of 0.89, and the N-back Working Memory Test, whose validity has been verified by experts and shows a reliability of 0.78, were used. The examinees in the test group received eighteen 90-minute sessions of training on cognitive-metacognitive strategies, while the examinees in the control group just received ordinary school training. Covariance analysis indicated a significant difference (p<0.001) between cognitive load and working memory grades between the two groups after training. Therefore, training on cognitive-metacognitive strategies as an intervention method can lead to reduced undesirable cognitive load and increased capacities of working memory in students with specific learning difficulties in reading.