Presented 265 undergraduates a list of words for study. Ss were then tested by a multiple-choice procedure in which number of wrong alternatives varied. In Exp. I, instructions to make frequency discriminations produced the same outcome as instructions to make recognition decisions. In Exp. II, recognition errors decreased 2% for each additional second of exposure duration during study and increased 6% for each additional wrong alternative present on the multiple-choice test. In Exp. III, errors in an absolute recognition test and in a 2-choice test corresponded in number to the errors made in absolute and 2-choice frequency-discrimination tests. Results are evaluated in conjunction with the hypothesis that frequency information is a dominant attribute in recognition decisions. (PsycINFO Database Record (c) 2006 APA, all rights reserved)