A study compared the writing abilities of children in a whole language classroom with those of children involved in a skills-based classroom. Subjects, 37 African-American students in two heterogeneous first-grade classrooms in a low socio-economic area school in the inner-city of a large Alabama city, completed a reading/writing assessment (a sample of the Stanford Achievement Writing Test for second-grade students) after 7 months of instruction. One group of subjects consisted of 19 students who had been instructed using a whole language curriculum; the other group consisted of 18 participants who had been instructed using a traditional curriculum emphasizing skill mastery. Writing samples were analyzed according to content and the ability to express thought. Total words and t-units (number of thoughts) were also compared between the groups. Results indicated that, on all tools of assessment, the whole language group scored significantly higher than the skill-oriented group. (Five tables of data are included; 18 references are attached.) (RS)