This study examined the effect of age on effectiveness of mnemonic instruction in enhancing reading abilities among grade three learners with dyslexia (LWD) in two public primary schools in Mpumalanga, South Africa. The Skinner’s reinforcement and the Information Processing theories were employed. A quasi-experimental design was used. Two schools were used as research sites, for intervention and another for control. A sample size of 43 learners was obtained in the two selected schools using purposive sampling technique. The tools used were the Bangor Dyslexia Test, pre- and post- tests, and a reading comprehension test. The results of reading ability indicate that the score of older learners (M=24.9; SD=10.3) was significantly higher than that of younger learners (M=15.8; SD=8.3), t (21) = 2.303, p = .032. The study recommended that the Department of Basic Education should revise the policy that reading is tested from grade one.