This study investigated the effects of recall rates (2:3-sec vs 2:5-sec vs 2:8-sec), length of between-trial intervals (none vs10Numerous investigations (see Paivio, in press) have indicated that concrete in comparison to abstract nouns function as superior stimulus cues for their associates in paired-associate learning (PAL). The phenomenon has been attributed to the capacity of concrete nouns to elicit sensory images which serve to mediate response retrieval. Paivio (in press) has theorized that imaginal mediation involves multiple coding and decoding stages (from words to images and back to words). Temporal factors of study-recall intervals may be assumed to interact with concreteness (C) factors. Yanney (1967) found such effects when length of learning intervals was varied. The present study investigated PAL as a function of Noun C, length of recall intervals, and length of between-trial intervals. Concrete word pairs are more likely to evoke imaginal than verbal mediators while the reverse is true of abstract word pairs (Paivio, Yuille, & Smythe, 1966). A test for reaction time (RT) by Paivio (1966) showed that image arousal was slower than verbal association to both concrete and abstract words; the concrete-abstract RT difference was much greater for imaginal than for verbal associative responses. Yuille & Paivio (1967) found that latency scores for discovery of verbal mediators did not differ between concrete noun pairs and abstract noun pairs. Paivio & Csapo (in press) investigated memory for pictures, concrete words, and abstract words at fast (3/l6-sec) and slow (Yz-sec) presentation rates. Measures were taken of memory span, free recall, recognition memory and serial learning. In general, memory for pictures was superior to concrete words and abstract words at the slow rate, but inferior at the fast rate. Memory for concrete words and abstract words was similar at the fast rate but concrete words were superior to abstract at the. slow rate. These findings on memory are congruent with the RT data. Since imaginal mediators should not be available for learning of concrete nouns at fast presentation rates but verbal mediators would be equally available it follows that PAL should not differ for concrete-concrete (C-C) , concrete-abstract (C-A) , abstractconcrete (A-C), and abstract-abstract (A-A) noun pairs at a fast rate. At slow rates, C-C and C-A pairs should be better recalled than A-C and A-A pairs since stimulus-elicited mediating imagery would enhance response retrieval.Yanney (1967) found that recall of abstract noun-nonsense syllable pairs learned at as: 5-sec presentation rate was superior to recall of concrete noun-syllables learned at a 3: 5-sec rate. It was expected that A-C and A-A PA learning at a slow rate would allow more time for verbal mediated retrieval and therefore superior recall than C-C and C-A pairs at a fast rate. The latter two S-R combinations would have less Psychon. Sci., 1968, Vol. 13 (6) time for verbal mediation than that given to A-C and A-A pairs and insufficient time ...