In five experiments with over 2,500 subjects, we examined the hypothesis that cognitive processing variables measuring breadth of declarative knowledge and information processing speed were related to learning outcomes on a paired-associates task. Experiments 1 and 2 compared recall with recognition tests, Experiment 3 assessed the effect of study-block size, Experiment 4 examined the effect of mnemonic strategy, and Experiment 5 tested the effect of mixing study times and presenting words versus nonsense syllable stimuli. Across all experiments, breadth of verbal knowledge was found to be a strong predictor of retention overall, and a strong predictor of increment in retention benefits due to increases in study time. Mnemonic strategy training improved retention but also served to enhance the relationship between knowledge and retention. Memory search speed also predicted retention, but primarily under conditions of high information flow, either as a result of short (5 seconds per pair) study or time-sharing pressure (mixed study-time blocks). High-Knowledge subjects and Fast Memory-Search subjects were also quicker at retrieving the answer, when they knew the answer; but High-Knowledge subjects took longer in retrieving an answer under conditions of uncertainty. Results are discussed in terms of a general model of associative learning in which encoding is viewed as a process of generating links by constructing elaborations of the terms studied. (Fourteen tables and 14 figures are provided. Appendixes discuss producing individual difference curves, and strategy-treatment group instructions for Experiment 4.) (Author)