The principal aim of strategy-focused instruction is to teach students strategies to control their writing processes and achieve quality writing. For this purpose, nine 4th grade Elementary School classes from three different schools (N = 215) were randomly allocated to two forms of strategy-focused program called cognitive selfregulation instruction (CSRI). The full-CSRI (experimental condition 1, n = 72) taught students a strategic approach to set appropriate product goals along with planning strategies. However, in the brief-CSRI (experimental condition 2, n = 69), the direct teaching of planning procedures was removed. These two experimental conditions were compared with a control condition (n = 74). We used a pre-test/posttest design and we also collected a maintenance writing performance 7 months after the intervention. Writing performance was holistically evaluated through readerbased measures made up of aspects related to structure, coherence, and quality. Only the full-CSRI condition wrote better compare-contrast texts than the control group in both the short term and at the maintenance timepoint. The study discusses the effects of the intervention on each measure and whether or not it is necessary to train process strategies.