The present experiments explored ways in which information encoded along different dimensions might be co-ordinated to determine memory performance. First, can an acoustic similarity decrement be offset by additional semantic encoding? Experiment I offers evidence for this compensatory interaction of codes. Second, does trade-off occur such that more information held in one code means less in another code? Third, are codes additive? Experiment II offers no support for the notion of a trade-off between encoding systems. However, when two codes were made available in the second study, recall was increased over the level achieved by either code alone.