Background: Middle school students use the information conveyed by morphemes (i.e., units of meaning such as prefixes, root words and suffixes) in different ways to support their literacy endeavours, suggesting the likelihood that morphological knowledge is multidimensional. This has important implications for assessment. Methods: The current study investigates the dimensionality of morphological knowledge considering the performance of 3,214 fifth through eighth graders on a range of morphological tasks (N = 14 across the project's development and 10 for dimensionality analyses) and items (N = 491) using multiple-group item response modelling. It then presents validation evidence related to performance of 1,140 fifth through eighth graders on a gamified, computer-adaptive, multidimensional assessment of morphological knowledge, which consists of seven morphological tasks and 181 items that make four morphological skills. Results: Results indicate morphological knowledge is multidimensional and best represented via a bifactor model of four skills as well as task-related variance. These skills are Skill 1: Morphological Awareness; Skill 2: Morphological-Syntactic Knowledge; Skill 3: Morphological-Semantic Knowledge; and Skill 4: Morphological-Orthographic/Phonological Knowledge. The assessment designed after this model, called Monster, PI, was shown to be both reliable and valid, with each morphological knowledge skill explaining unique variance in standardised reading vocabulary. Conclusions: Findings suggest that morphological skills play unique roles in language and literacy outcomes. This indicates the importance of conceptualising and assessing morphological knowledge as multidimensional. Implications for theory, research, policy and practice are considered. [This paper was published in "Journal of Research in Reading" v44 n1 p70-89 2021 (EJ1283025).]