Background: In the field of intelligence research, it has been shown that some people conceive intelligence as a fixed trait that cannot be changed (entity beliefs), whereas others conceive it as a malleable trait that can be developed (incremental beliefs). What about writing? Do people hold similar implicit theories about the nature of their writing ability? Furthermore, are these beliefs likely to influence students' response to a writing intervention? Aims: We aimed to develop a scale to measure students' implicit theories of writing (pilot study) and to test whether these beliefs influence strategy-instruction effectiveness (intervention study). Sample: In the pilot and intervention studies participated, respectively, 128 and 192 students (Grades 5-6). Method: Based on existing instruments that measure self-theories of intelligence, we developed the Implicit Theories of Writing (ITW) scale that was tested with the pilot sample. In the intervention study, 109 students received planning instruction based on the self-regulated strategy development model, whereas 83 students received standard writing instruction. Students were evaluated before, in the middle, and after instruction. Results: ITW's validity was supported by piloting results and their successful cross-validation in the intervention study. In this, intervention students wrote longer and better texts than control students. Moreover, latent growth curve modelling showed that the more the intervention students conceived writing as a malleable skill, the more the quality of their texts improved. Conclusion: This research is of educational relevance because it provides a measure to evaluate students' implicit theories of writing and shows their impact on response to intervention.