We used two measures to assess whether students could identify units of meaning, both in terms of breaking words down and also in terms of connecting to larger words. We used a task titled Odd Man Out,where students were given three words and had to identify the word that did not belong. For example, when looking at estimate, classmate, and roommate, the morpheme mate in classmate and roommate represents the meaning of ‘a person’. Estimate is the odd man out: it is the example where the morphemes do not overlap in terms of meaning. This task forces students to think about how the written language conveys meaning. Our second measure, which we call Meaning Puzzles, asks students to identify the word part that is most helpful in determining the meaning of that word, which in this case can be a morpheme or a morphologically related word that is contained in the larger word. Here,students need to look beyond overlap in spelling to figure out the link to meaning. For example, accusatory has spelling overlap with accurate, accuse, cushion, and custom, but only accuse overlaps in terms of meaning. Hence, a student who knows they can use accuse to figure out accusatory rather than using accurate is more likely to figure out the word’s meaning and apply that meaning to their literacy endeavors. It is important to note that while we had developed other tasks to identify units of meaning in words, these tasks worked best.In terms of instruction, this skill highlights that we need to help our students think about morphological overlap between words. We need to constantly challenge students to think about how the word’s form and sound conveys links to other words in their morphological family. For example, thinking about astronomy’s relationship to astronaut,and then about the units like astr=star,will build a strong foundation of understanding how units of meaning are put together to convey meaning or even broken down to figure out meaning. We have observed instructional work where students eagerly look up word origins to find overlap, create word webs to identify morphological word families, and find imposters where tricky spellings look like morphemes but are not (see Goodwin & Perkins, 2015 for examples).