State standards include fluent use of standard computational algorithms. However, learning and using them require conceptual understanding of numbers and operations. Previous research using the concrete--representational--abstract (CRA) sequence has been effective in teaching computational algorithms to students at risk of and students with learning disabilities by emphasizing conceptual understanding. However, all the research was face-to-face and few captured the impact of the intervention on students' conceptual understanding. The current study occurred during the COVID-19 pandemic, so researchers modified CRA for remote instruction. This study investigated the effects of modified CRA on the fluency and accuracy of three sixth-grade students with learning disabilities. They also included an assessment of conceptual understanding. The researchers used a multiple probe across participants design, demonstrating a functional relation between CRA and students' fluency and accuracy. Researchers assessed conceptual understanding with informal measures that required application of their conceptual understanding. The results and implications are discussed.