Attentional templates can be represented in visual working memory (VWM) when the target varies from trial-to-trial and can be represented in long-term memory (LTM) when the target is consistent during trial runs. Given that attentional templates can be represented in either VWM or LTM, are there any differences in how these representations impact visual search when targets are consistent compared with varying? The current study tested the consistent template hypothesis, which predicts faster performance with a consistent target compared with a varying target. Experiment 1 examined whether consistent targets could lead to consistent templates that would improve template establishment, guidance, and/or comparison of the template to search items. Search response time was faster for consistent targets, and consistent targets produced faster comparison processes, but not more efficient guidance. Experiment 2 examined the consistent template restoration hypothesis, which predicts faster template establishment and comparison processes for a previously encountered consistent target. Experiment 2 replicated the consistent template hypothesis and supported the consistent template restoration hypothesis. These studies demonstrate that although attentional guidance is similar with varying and consistent attentional templates, consistent templates improve search performance by speeding template establishment and comparison processes.