This research compared two methods of teaching writing in a foreign language, applied over a period of ten weeks. The sample was composed of two groups of German secondary school students, who were studying Spanish as a foreign language at the A1 level. This study, which had a pretest/posttest/delayed posttest design, measured the ability of the participants in the experimental group (EG) and the control group (CG) to produce written texts with an acceptable level of linguistic correction, coherence, cohesion, and adequacy. The results showed that the texts written by the EG (32 subjects) improved in all four parameters analyzed. In contrast, there was no improvement in the written production of the CG (32 subjects). Furthermore, a delayed posttest, administered three weeks later at the end of the didactic treatment, revealed that the declarative knowledge, acquired by the EG subjects during the four semi-controlled tasks in the didactic treatment, had become procedural knowledge. This higher level of writing competence was evidenced in the posttest in relation to all four discourse parameters.