This study of 31 college students employed the startle eye-blink modification (SEM) technique to index both early and later stages of attentional processing during a memory-load version of the Continuous Performance Test (CPT). Participants viewed a series of digits and pressed a button after the digit 7 of each 3–7 sequence. A startling noise burst was presented either 120 or 1,200 ms following three preselected prepulses: target (3), nontarget (non-3 and non-7 digits), or target plus distractor (3 and simultaneous tone distractor). Greater startle inhibition occurred 120 ms following target and target-plus-distractor prepulses compared with nontargets, indicating early selective attention. No difference was observed between SEM during target and target-plus-distractor prepulses, suggesting the distractor was effectively ignored. At 1,200 ms, the three prepulse types produced nondifferential inhibition, suggesting that modality-specific selective attention occurs in anticipation of the presentation of the next CPT prepulse. These findings indicate that SEM distinguishes between different early selective attention and later anticipatory attention subprocesses underlying the CPT.