When paying attention to a feature (e.g., red), no attentional advantage is gained in perceiving items with this feature in very brief displays. Therefore, feature-based attention seems to be slow. In previous feature-based attention studies, attention has often been measured as the difference in performance in a secondary task. In our recent work on Boolean map theory (Huang & Pashler, 2007), we distinguished between 2 concepts that are often conflated with the term "attention", namely the "selection of information" from stimulus and the following "processing optimization" (i.e., attentional advantage) of the selected stimulus. Attention, as examined in previous feature-based attention studies, only fits the definition of processing optimization, but does not fit the definition of "selection of information". Therefore, it is open to question whether feature-based attention, when defined as "selection", is fast or slow. In this study, I systematically measured the speed of feature-based attention in relation to both definitions. Attention was found to be slow ([image omitted]100 ms) in terms of processing optimization (i.e., attentional advantage) but fast in terms of the selection of information (less than 50 ms). These results support the view that feature-based attention works by creating a spatial representation (i.e., a "Boolean map"; Huang & Pashler, 2007) for the stimulus of a feature and a "processing optimization" of the visual information residing in the region of this spatial representation. (Contains 2 footnotes and 6 figures.)