In order to test the hypothesis that bizarre imagery can be an effective mnemonic aid with delayed testing, a context of mixed materials, and an adequate stimulus presentation pace, a study examined 40 undergraduates who were randomly presented with three paired-associate lists (normal, bizarre, and mixed). Within each list the sentences consisted of stimulus-response pairs of high-imagery nouns. Twenty subjects participated in an immediate cued recall condition for each list while the remaining 20 were tested after a 1-week retention interval. Subjects were further subdivided depending upon whether the presentation pace was forced or not. Results indicated that bizarre imagery would seem to be an effective mnemonic aid on immediate recall when materials are mixed and when the pacing is self-controlled. However, if long-term retention is required, the learning context is the important factor on performance. (A table of the mean number of correct terms recalled as a function of context, delay, and presentation pace is included, and 28 references are appended.) (MM)