A processing-capacity notion of attention was defined in terms of increased reaction times (RTs) in a probe task at specific times during the performance of a primary task The primary task involved a same-different response to two letters presented successively. The probe task was either a Bonders type a or type c reaction to a burst of white noise, which could occur at any one of 8 times relative to the events on the letter-matching task. The study was specif ically designed to investigate the encoding of the f irst letter , which Posner and Boies (1971) concluded required no processing capacity. Here, displaying the first letter for 15 msec, and following it with a visual mask did result in lengthened RTs to probes presented simultaneously with the first letter. Increasing the difficulty of the probe task did not result in greater interference from the primary task. Data are discussed in terms of processes requiring capacity.