An earlier study had concluded that (a) facilitation of short-term recall for auditory digit sequences by increases in digit duration were due to increased rehearsal time available under these conditions, and that this extra rehearsal time was more facilitative at longer rather than at shorter retention intervals; and (b) efficient encoding techniques improved recall regardless of retention interval length. The present study not only confirmed these conclusions for a situation where digit sequences received both a visual and an auditory encoding but also determined that the digit duration which is critical in terms of allowing significantly more rehearsal time is between .5 and 1 sec.Recent studies (Haber & Nathanson, 1969;Miscik, Smith, Hamm, Deffenbacher, & Brown, 1972 ;Sitterly, 1968) have shown that improvements in short-term retention of digit strings can be accomplished by increases in digit duration as well as interdigit interval, the latter effect having already been well established. In addition, Miscik et al. found that digit duration interacted with retention interval in that the facilitatory effect of longer durations was greater at longer retention intervals than at shorter ones. Perhaps as Miscik et al. suggested, extra rehearsal time associated with longer digit durations made these items more resistant to forgetting at longer retention intervals.Not surprisingly, Miscik et al. (1972) also found that more efficient encoding techniques (chunking) improved short-term retention of auditorily presented digits. Somewhat less expected, however, was the finding that chunking the digits was consistently superior to not chunking them at all digit durations and at all retention intervals. These two results suggested that (a) facilitation of recall by longer durations apparently cannot be explained in terms of longer durations allowing more time for chunking to operate, as 5s were able to chunk effectively at even the briefest duration (.5 sec); and (6) chunking either increased resistance to forgetting during stimulus presentation or permitted more effective retrieval regardless of retention interval.The present experiment attempted to determine whether the Miscik et al. (1972) findings also hold for a situation where digit strings receive a visual as well as an auditory encoding. Digit strings are presented visually with 5s required to group them aloud. If an increase in digit duration from .5 sec. to 1.0 sec. in the Miscik et al. study was critical in terms of extra rehearsal time, it would seem of interest to know whether parametric variations in duration less than .5 sec. are also critical to rehearsal 1 Special thanks are extended to Jerry Jarombek for his assistance in data collection and analysis and to Evan Brown for his helpful comments and suggestions.