Among scholars of the Ming-Qing imperial examinations, not much attention has been paid to the subject of the eight-legged essay, much less to the issues related to xiaoti (small version of the eight-legged essay test). Indeed, the existing scholarship on xiaoti is at best fragmented; at worst, it is marked by instances of misunderstanding and conflicting accounts. In this paper I propose to re-open the study of xiaoti. I discuss the nature and content of xiaoti, as well as that of jiedati (test pm fragmented texts taken from the classics). In addition, I analyze the factors which gave rise to the appearance of xiaoti and jiedati, and explain their advantages in preventing niti (guessing the test questions beforehand) and plagiarism, as well as in making the grading process easier. In particular, I focus on the following questions: Was it necessary that the tests of xiaoti contained shorter statements than the normal dati (big vesion)? As to the tests which contained merely a single-sentence statement (danjuti), did they belong to the category of xiaoti, or that of dati? Finally, did the tests of xiaoti actually display the tendency of having shortened statements? My basic contention is that the naming of dati and xiaoti actually originated from the ranking of different levels of examinations-provincial, metropolitan, and county examinations-where the tests took place. In other words, as far as the content of the test was concerned, the distinction between dati and xiaoti was not determined by its length, but by its comprehensiveness and complexity. As to the first appearance of xiaoti, I suggest that it be dated to the Mind Chenghua reign (1465-1487), during which more and more of the county examinations were inclined to adopt the formulation of xiaoti. From then on and through the Hongzhi reign (1488-1505), notable experts on xiaoti began to arrive on the scene. As the popularity of xiaoti reached its climax in the Wanli reign (1573-1620), it also facilitated the first appearance of jiedati.