This study investigates the role of aging on story comprehension and r
ecall and also the relationships of these variables to working memory
capacity. The research involved two age groups, younger subjects (mean
age 21 years) and older subjects (mean age 64 years). Comprehension w
as defined as th construction of Cognitive-Semantic Structures (c-s s)
resulting mainly from the selection and integration of units into ord
inal structures implying inferential reasoning. Two reading conditions
were distinguished non-directed reading (with classical instructions)
and directed reading involving the verification of statements leading
the subject to the construction of c-s s. Results show that reading c
ondition affected reading time, but did not affect comprehension or re
call scores. No interaction was observed between reading condition and
age. Consequently, in each age group, new sub groups were formed on t
he basis of the comprehension scores. Our hypothesis was confirmed; co
mprehension and recall scores were lower in the older group than in th
e younger group. The relationship between recall and comprehension was
modulated as a function of age. In young subjects, recall increased i
n a systematic way as a function of comprehension scores. In older sub
jects, recall was higher for subjects who achieved better comprehensio
n scores, but the recall improvement due to better comprehension was l
ess than for young subjects. This result is interpreted in terms of cu
mulated deficits concerning comprehension, storage and retrieval proce
sses involved in recall. The capacity of working memory was measured w
ith a reading span test (first version). No difference was observed be
tween mean reading span for younger and older subjects. For the younge
r subjects, reading span did not correlate with comprehension or recal
l scores. For the older subjects, reading span did not correlate with
comprehension scores but correlated with recall scores. These results
are discussed, taking into consideration the appropriateness of the re
ading span. test first version to measure the capacity of working memo
ry on both processing and storage functions.