The present study is a part of serial studies of the testing effect. We assume "the motivating power" as the most important element of the effect of test. Considering from the viewpoint of the elements of this motivating power, pupils in advance may be supposed an important source of the motivating power. Because, it is possible to think that when pupils are instructed to be given a test, most of them would be drived to review that materials in order to get good score or to avoid poor score. We attempted, therefore, four experiments to know the effect of annoucement of the coming test. The learning materials consisted of a number of dictations of Kanji (chinese character), spellings of English words and some arithmetic problems all of which were to be used in regular school learning situations. Experiment I ane II were made with the primary school pupils as subjects, in the so-called controlled-group method, that is, on the basis of the pretest result, the Ss were divided into two equivalent groups : the control and experimental groups. Only the experimental group was informed that on the following day they would take a test of the previously learned materials, while the control group was not instructed at all. Then, both groups sat for the same critical test. The difference of scores obtained from the groups made it possible to analyse the effect of announcement of the coming test. But the difference was statistically not fully significant. (see Table 1, 2 and 5). Experiment III and IV, modified in respect to two or three procedures by reviewing previous experiments, were given to secondary school pupils with the rotation method. In the experiment III, the two comparisons of both group's scores showed a superiority of the experimental group with a fully significant difference. (see Table 3 and 5). In this experiment IV, the experimental groups had again more advantage than the control groups, but there was not fully significant difference between both groups. (see Table 4 and 5) Although these experimental results were individual not statistically significant, their combined effect supports the hypothesis that the annoucement of coming test had a motivation power that have the positive effect on performance. It should be pointed out that a further investigation of the effect of annoucement of the coming test under the various conditions of learning materials, abilities, and sex, etc. ramains an important problem.