Purpose: We performed a meta-analysis to examine if children with dyslexia experience deficits in morphological awareness (MA) and if the effect sizes are influenced by different moderators (age, aspect of MA measured, type of MA task, language, modality input, semantic knowledge, and selection criteria). Method: We reviewed 40 studies published in English between January 1990 and August 2021, representing a total of 5,018 participants (age range = 5.1-13.9 years). Studies with adults or English language learners were excluded. There were 49 independent samples in the chronological-age (CA) -- dyslexia (DYS) comparison and 18 independent samples in the comparison between DYS and reading--level (RL) controls. Results: A random-effects model analysis revealed a large effect size for the CA--DYS comparison (g = 1.11) and a non-significant effect size for the RL--DYS comparison (g = -0.08). Age was the only significant moderator of the effect sizes. Conclusion: These findings suggest, first, that individuals with dyslexia experience significant difficulties in MA and second, that the effect sizes are as large as those reported for phonological awareness, rapid automatized naming, and orthographic knowledge. The lack of a significant RL-DYS difference indicates that MA is not a core causal feature of dyslexia.