This review examines the literature on assessing borderline personality disorder (BPD) using the MMPI and the Rorschach test. Despite the extensive use of these instruments in the assessment of BPD, the degree to which they are successful in identifying the disorder remains unclear. Methodological difficulties in this literature are discussed, including diagnostic and sample heterogeneity, base rate issues, systematic differences in selection criteria among studies, and, for the Rorschach, differing outcome measures. Both MMPI and Rorschach findings suggest a general trend for borderline groups to produce relatively unique findings. Their usefulness in yielding accurate individual diagnoses of BPD, however, has not yet been established. Furthermore, no supportive evidence was found for the commonly held hypothesis that BPD subjects show more impairment on unstructured measures than on objective measures.