A substantial body of research has demonstrated the efficacy of early writing interventions that target skills and composing processes; however, much less is known about the effectiveness of classroom writing instruction outside of the context of an intervention. The goal of this study was to investigate if writing instruction and student practice predict first-grade writing achievement and if the relations between writing instruction, practice and achievement depend on student factors. Assessments of students' spelling, handwriting, vocabulary, and reading were collected in the fall of first grade, and norm-referenced and researcher-designed writing tasks were administered in the spring (N=391). During the school year, four full-day observations of classroom instruction and student writing practice were conducted in 50 classrooms. The effects of writing instruction and student writing practice on spring writing achievement were analyzed using two-level, fixed-effects hierarchical linear models. Composing instruction was negatively related to contextualized spelling, but no other main effects of instruction were found. One type of writing practice, generative writing, was positively related to all three measures of writing achievement. Interactions were also found between student gender, minority status and multiple types of writing instruction and practice. These results point to the potential benefit of generative writing practice and indicate that efforts to differentiate instruction and practice may be beneficial for students. Additionally, the findings raise doubts about the effectiveness of current writing instruction. [This paper was published in "Contemporary Educational Psychology."]