In this study the effect of a writing strategy for L2-writing is tested experimentally. 10 th grade students (N = 67) took part in a writing experiment in writing German as a second language. In the experimental condition students were taught to pay attention to pre-writing activities (think and organize), find the right words in German, and post-writing activities (evaluate and revise). Students in both the experimental and control condition completed two texts as pre-and post-test. All texts were rated holistically as well as with an analytic scoring scheme. Results show that in the experimental condition students wrote (on average) better texts than in the control condition, when rated holistically. For the analytic ratings improvement was statistically significant for the categories Content, Vocabulary and Conventions. For other categories (Syntax, Grammar (Verb and Case), Spelling and Punctuation) no effect of the experimental program could be shown.