Predictive arrow cues, as used in the classic “Posner paradigm”, that were long thought to engage and isolate voluntary attention, may in fact trigger a strong interaction between voluntary and involuntary attention (Ristic & Kingstone, 2006). This interaction produces an orienting effect that exceeds both the effects of involuntary and voluntary attention alone, and the additive combination of involuntary and voluntary orienting. The present study shows that nonpredictive peripheral cues—understood to engage and isolate involuntary attention—if made predictive, result in enhanced orienting effects similar to predictive arrows. The important contribution of these data is that they suggest an “interaction principle”: If attention cues can elicit reliable involuntary orienting, then when they are made spatially predictive, the resulting attention effect will be greater than the sum of involuntary and voluntary orienting alone.