The present study used a video-based experimental design to investigate the influence of visible roadwork activity on speed preferences at work zones. Four videos from real work zones in Norway were used. Two roadwork areas were filmed at two moments — with and without visible roadwork activity. A total of 815 drivers watched two videos and answered a questionnaire online. Participants stated speed preferences for both videos and then evaluated the influence of 17 common work zone elements on their speed choice. The results showed lower preferred speeds for the videos with visible roadwork activity. The elements considered by drivers to influence their speed included speed regulation (e.g. speed limit signs), transient motives (e.g. time pressure), flow pressure (e.g. speed of the rear driver) and situational conditions (e.g. road situation). Results from the regression analysis showed that visible roadwork activity was the strongest predictor of lower preferred speeds. Speed regulation and transient motives were also significant predictors of preferred speeds. Those who considered speed regulation more important to influence their speed choice at work zones were likely to state lower preferred speeds, while those who considered transient motives as more important were likely to state higher preferred speeds. These findings support the hypothesis that visible roadwork activity is an important factor for lower speed preferences at work zones. The main practical implication is that road authorities should consider various countermeasures to achieve safe driving speeds at work zones, not only roadworks warning signs.