This study examines the effectiveness of different visual attention prompt methods in directing drivers’ gaze toward critical traffic targets after short-term training. Using a Unity-based driving simulator and eye-tracking technology, five methods—Point, Arrow, Blur, Dusk, and a Modified Attention Funnel—were tested. Eye-tracking data showed that visual prompts improved detection efficiency, with effectiveness varying by driver proficiency. The results indicated that Point, Arrow and Dusk positively influenced drivers’ efficiency in detecting critical objects in a short period. Additionally, driver proficiency significantly moderates the effectiveness of these methods: the Arrow method was most effective for experienced drivers, whereas the Dusk method provided better guidance for novice drivers.