After taking a personality test, subjects reported their reactions to being tested. Reactions were diverse, even in the same subject. Free responses to 10 questions were coded into 16 categories within five broad groups. Desire for information about the test and about self, and criticism of testing were very prevalent; criticisms of self and apprehensions were less frequent. Three studies yielded rather similar results. Inventories, however, elicit somewhat different patterns of reactions from other tests. These diverse reactions may underlie response sets and, consequently, impair the quality of personality measurements. Appropriate increases in the structuration of our tests seem called for. (Author)