Subjects read aloud words presented once at the rate of one per second. A perceptual identification task, involving 30-or 50-msec presentations, followed. Some of the words presented for identification had been read previously; others were new. After each presentation, in addition to identifying the word, the subjects judged its duration. The data indicate that a single presentation of a word affects its later perception, as revealed by enhanced perceptual identification, longer duration judgments, and better temporal discrimination. A second experiment showed that a single presentation influenced duration judgments even when identification was not required. The final experiment addressed the issue of what is preserved in memory from a prior presentation. The results from the three experiments indicate that duration judgments provide a valuable dependent measure of memory in the perceptual identification task and support the misattribution hypothesis: A prior presentation enhances perceptual identification, and this increase in relative perceptual fluency is incorrectly attributed to a longer presentation duration.Many who have traveled in a foreign country in which an unfamiliar language is spoken have commented that the native speakers seem to talk much faster than people back home. It is as though lack of familiarity with a language makes the words rush by, whereas familiarity makes the words hang in the air longer than they actually do.1 This subjective impression that familiarity affects judgments of duration has been supported experimentally as well. For example, Devane (1974), Warm, Greenberg, and Dube (1964), and Warm and McCray (1969 have reported that the durations of presentation of familiar words, as defined by frequency of usage in Thorndike and Lorge (1944), were judged to be longer than those of unfamiliar words in tachistoscopic identification tasks.A number of studies (e.g., Feustel, Shiffrin, & Salasoo, 1983;Jacoby, 1983aJacoby, , 1983bJacoby, , 1984Jacoby & Dallas, 1981;Jacoby & Witherspoon, 1982;Morton, 1979;Murrell & Morton, 1974;Witherspoon, 1983) have shown that one prior presentation of an item enhances subsequent perceptual identification under tachistoscopic conditions. In the experiments reported by Jacoby and Dallas (1981), the subjects first read aloud a list of words, each word being presented once on the screen at a rate of about one per second. A second list of words was then presented, each word now being shown for a brief constant duration of about 35 msec. The second (test) list contained some words from the original (study) list and also new words