THIS STUDY WAS MADE TO COMPARE THE EFFECTS OF STIMULATION AND NONSTIMULATION ON RECALL OF WORDS FOLLOWING TIME-DELAY PERIODS. THE SUBJECTS (670 AIRMEN) WERE TRAINED WITH AN EXAMPLE WORD LIST AND TWO WORD LISTS CONTAINING FIVE OF THE SECONDARY WORDS ASSOCIATED WITH RUSSELL-JENKINS STIMULUS WORDS. AFTER TIME DELAYS OF 2 MINUTES, 20 MINUTES, 2 DAYS, OR 28 DAYS, 24 GROUPS OF SUBJECTS WERE TESTED ON RECALL OF THE SECOND LIST. ON THE RECALL TEST, HALF OF THE SUBJECTS WERE GIVEN THE FIVE STIMULUS WORDS CORRESPONDING TO THE SECOND WORD LIST, AND HALF WERE NOT. THOSE SUBJECTS TESTED AFTER 28-DAY TIME DELAYS STILL SHOWED STRONG EVIDENCE OF THE WORD LIST TRAINING, BUT THE PERFORMANCE WAS MORE LIKE THAT OF FREE ASSOCIATING SUBJECTS THAN THAT OF SHORTER RETENTION SUBJECTS. IN THE NONSTIMULATED GROUP, INTRUSIONS FROM UNIDENTIFIED SOURCES WERE MORE NUMEROUS WHEN THE TIME PERIOD WAS LONGER. AS RECALLS OF WORDS ON LIST TWO DECREASED WITH TIME, INTRUSIONS OF WORDS FROM THE EXAMPLE LIST AND FROM THE FIRST TRAINING LIST WERE MORE NUMEROUS WHERE CULTURAL PROBABILITIES SUGGESTED THIS OUGHT TO HAPPEN. (AL)