We investigated whether congruent action while reading informational texts about physics improves science, technology, engineering, and math (STEM) vocabulary learning and reading comprehension. Second- and third-grade Spanish-speaking children (N = 216) were randomly assigned to one of five conditions implementing forms of Embodied-Action or Read-Only no-action conditions. In the three Embodied-Action conditions: (a) children individually read and moved images on an iPad screen to simulate the meaning of the sentences; or (b) children read and moved images and an intelligent tutoring system updated a vocabulary page depending on how often children tapped on target words in the text; or (c) the experimenter and children both read aloud and used whole-body pantomime to mimic the actions presented in the text. In the two Read-Only no-action conditions: (d) children individually read the text on the iPad, but did not perform any movement of the images on the screen; or (e) the experimenter and children read the text aloud together but did not perform any pantomime. Our results show that both types of Embodied Action (using screen or pantomime) significantly enhanced informational text comprehension and STEM-related vocabulary learning relative to the Read-Only no-action conditions. Our findings support the effectiveness of incorporating embodied action, whether or not technology is used. Teachers can easily implement action-based vocabulary and reading comprehension strategies in their classrooms.