The article presents a series of experiments that challenge the traditional feature integration model of visual search, which posits serial, self-terminating search for conjunctions of features. The authors, Jeremy M. Wolfe, Kyle R. Cave, and Susan L. Franzel, from the Department of Brain and Cognitive Sciences at MIT, conducted experiments where subjects searched for targets defined by conjunctions of color and form, color and orientation, or color and size. They found that for many unpracticed subjects, the reaction time (RT) × set size functions were too shallow to be consistent with serial search. Searches for triple conjunctions (color × orientation × size) were particularly efficient and independent of set size, which contradicts the feature integration model. The authors propose a guided search model, similar to Hoffman’s two-stage model, where parallel processes use information about simple features to guide attention in the search for conjunctions. This model explains why triple conjunctions are found more efficiently than standard conjunctions, as three parallel processes can guide attention more effectively than two. The guided search model also accounts for the nonlinear slopes observed in some conjunction searches, which may be due to the density of items in the display or other factors like stimulus salience. The article concludes with three main points: 1. The results from naive, unpracticed subjects searching for conjunctions of color and form, color and orientation, and color and size are inconsistent with serial, self-terminating search. 2. Searches for triple conjunctions are easier than for simple conjunctions, which is not predicted by the standard feature integration model. 3. A modification of the feature integration model to allow parallel processes to guide serial search can explain these and other problematical results.The article presents a series of experiments that challenge the traditional feature integration model of visual search, which posits serial, self-terminating search for conjunctions of features. The authors, Jeremy M. Wolfe, Kyle R. Cave, and Susan L. Franzel, from the Department of Brain and Cognitive Sciences at MIT, conducted experiments where subjects searched for targets defined by conjunctions of color and form, color and orientation, or color and size. They found that for many unpracticed subjects, the reaction time (RT) × set size functions were too shallow to be consistent with serial search. Searches for triple conjunctions (color × orientation × size) were particularly efficient and independent of set size, which contradicts the feature integration model. The authors propose a guided search model, similar to Hoffman’s two-stage model, where parallel processes use information about simple features to guide attention in the search for conjunctions. This model explains why triple conjunctions are found more efficiently than standard conjunctions, as three parallel processes can guide attention more effectively than two. The guided search model also accounts for the nonlinear slopes observed in some conjunction searches, which may be due to the density of items in the display or other factors like stimulus salience. The article concludes with three main points: 1. The results from naive, unpracticed subjects searching for conjunctions of color and form, color and orientation, and color and size are inconsistent with serial, self-terminating search. 2. Searches for triple conjunctions are easier than for simple conjunctions, which is not predicted by the standard feature integration model. 3. A modification of the feature integration model to allow parallel processes to guide serial search can explain these and other problematical results.