The current study aimed to investigate the effectiveness of an online dynamic test in reading and writing, differentiating in typically developing children (n = 47) and children diagnosed with dyslexia (n = 30) aged between nine and twelve years. In doing so, it was analysed whether visual working memory, auditory working memory, inhibition, cognitive flexibility, and reading self-concept were related to the outcomes of the online dynamic test. The study followed a pretest-training-posttest design with two conditions: experimental (n = 41), who received training between the pretest and posttest, and control (n = 37), who received training after the posttest. Results showed that typically developing children and children diagnosed with dyslexia in both conditions could improve their reading and writing accuracy scores, while the training in prosodic awareness might have tapped into children's potential for learning. Moreover, results revealed that in children diagnosed with dyslexia, training in the domain of writing competence could compensate for cognitive flexibility. However, training was not found to compensate for reading self-concept in children diagnosed with dyslexia.