Normal personality traits, as measured by the Zuckerman-Kuhlman Personality Questionnaire (ZKPQ), predicted some personality disorders in a sample of healthy volunteers. Whether these predictions could be more pronounced in patients with personality disorders remains unknown. We administered the ZKPQ and the Parker Personality Measure (PERM), which describes the functioning styles of personality disorder, in 134 patients with a range of personality disorders and in 268 age-, gender- and education level-matched healthy volunteers. Cluster A patients scored lowest on Sociability, cluster B highest on Impulsive Sensation Seeking and Aggression-Hostility, cluster C1 (Avoidant and Dependent types) highest on Neuroticism-Anxiety, and cluster C2 (Obsessive-Compulsive type) highest on Activity. Most of the predictors were consistent across both the healthy and patient groups. The variances that accounted for predicting most PERM styles by the ZKPQ traits in the patient group were higher than those in the healthy group. Our results showed that the ZKPQ traits could specifically predict the PERM styles in both healthy subjects and personality-disorder patients. This result was more pronounced in the latter group. The most powerful predictions were obtained for Antisocial, Dependent, Borderline and Avoidant styles, and the weakest for the Schizotypal and Schizoid styles in the patient group.