This study aimed to describe the writing activities that test takers engage in when responding to the writing tasks in the "TOEFL iBT"[superscript R] test and to examine the effects of task type and test-taker English language proficiency (ELP) and keyboarding skills on the frequency and distribution of these activities. Each of 22 test takers with different levels of ELP (low vs. high) and keyboarding skills (low vs. high) responded to 2 TOEFL iBT writing tasks (independent and integrated) on the computer. Each participant then provided stimulated recalls about the writing activities they used when performing each writing task. Stimulated recalls were coded and the results were compared across tasks and test-taker groups. The findings indicated that the participants engaged in various construct-relevant activities, such as interacting with the writing task and resources, planning, generating, evaluating, and revising. Additionally, test takers' writing activities varied significantly across tasks and to a lesser extent across test-taker groups. Participants' writing activities varied most across writing tasks and, to a lesser extent, across English proficiency groups. Low keyboarding skills seem to have affected mainly activities on the independent writing task. To better understand the role of keyboarding skills in performance on the TOEFL iBT writing tasks and to address the test's extrapolation inference, future studies need to compare the writing performance of test takers with different levels of second language (L2) proficiency and keyboarding skills in test and nontest settings.