A brief experiment was designed to examine cognitive flexibility practice embedded in beginning phonics instruction for kindergarteners with limited early literacy learning. Previously tested phonics content included single- and high-frequency two-letter grapheme-phoneme correspondences (GPCs), introduced at a rate of 2-4 correspondences per week. Children entered with minimal alphabet knowledge and were randomly assigned within classrooms, stratified by English Learner (EL) status, to one of two conditions delivered individually over six weeks: Plain explicit phonics (n = 35) or Flex instruction (n = 33) which covered the same Plain phonics content but with teaching tasks for practice switching letter or word dimensions. Results showed that kindergarteners in the Plain condition made significantly greater gains on tasks of writing taught letter-sound correspondences and spelling. Findings inform a rate for introducing letter-sound correspondences and learning of mixed-size GPCs. The Plain explicit phonics focus on initial accuracy had benefits for encoding taught letter correspondences. Findings support future research on effective tasks to develop reading-related cognitive flexibility in beginners, the optimal timing of this practice, and whether it benefits in particular children most at risk for acquiring foundational alphabetic knowledge. [This paper will be published in "Reading and Writing: An Interdisciplinary Journal."]