This study aimed to investigate how Grade 1-2 English language learners (ELLs) differ in their performance on a writing test in two test modes: paper and online. Participants were 139 ELLs in the United States. They completed three writing tasks, representing three test modes: (1) a paper in which students completed their writing using a paper-and-pencil format, (2) online-handwriting in which students viewed the prompt online and handwrote responses, and (3) online-keyboarding in which learners viewed the prompt online and keyboarded responses. Up to 20 students were tested during each 90-minute session, during which researchers observed two or three students to examine any differences in their writing behaviors across tasks. Immediately after students completed the tasks, researchers asked them about their preferred test modes. Students' writings were analyzed quantitatively and qualitatively in terms of task completion, fluency, and complexity. Interview data were analyzed qualitatively. Findings suggest the handwriting task is the optimal mode for students in Grades 1-2. Yet the majority of ELLs observed said they preferred the keyboarding task over the handwriting tasks. Study results provide practical implications for writing test administration for K-12 ELLs.