A sample of 15 eighth-grade Hispanic students in a bilingual classroom were used for a descriptive analysis of students' writing samples to compare their growth between pre- and post- writing samples in Spanish and English. This was accomplished by juxtaposing English and Spanish pre- and post-tests using the same holistic rubric developed by the school district for reevaluating the writing samples. Findings suggest that development of literacy skills for bilinguals follows a normal progression. The results also suggest that a holistic technique intended not only to complement a process approach to writing, but to focus on a global impression of students' ability to communicate ideas effectively, is not sufficient to ensure fairness in evaluation. It is noted that poorly developed holistic rubrics, which do not detect important gains or differences in writing levels, frequently mask academic potential in bilingual students and are generally biased in favor of English. To guard against these inherent biases, school districts should use more caution in assessing the academic skills of bilingual students, ensuring that they are given credit for knowledge and acquisition of skills even if these skills are not in English. (GLR)