The aim of the research in question was to investigate the efficiency of various classes of distractors used in multiple-choice vocabulary question testing. Much of the quality of multiple-choice questions relies on the extent to which incorrect choices tempt the less proficient student. The study selected 8 classes of distractors and attempted to discover whether their effectiveness could be quantified and a rank order of preference could be established for population samples of different levels of English proficiency. The study, involving 1200 Arabic-speaking students of English, found a consistent order of preference and some puzzling questions. It was found that a definite hierarchy of distractor efficiency exists in terms of discrimination and, particularly, potency. All groups demonstrated a preference for distractor types, but there was an uneven agreement in discrimination. In addition, the most popular distractors were not necessarily the most discriminating. Differences between populations of varied proficiency were greater than between groups from different geographical locations. (Author/KM)