This study used a desktop virtual environment (VE) of 15 large-scale residential streets to test the effects of environmental mystery and surprise on response. In theory, mystery and surprise should increase interest and visual appeal. For each VE, participants walked through an approach street and turned right onto a post-turn street. We designed the streets to have low mystery (straight) or higher mystery (curved); and we designed the street-pairs to have low surprise (no difference between them), medium surprise (moderate difference between them), or higher surprise (higher difference between them). A total of 37 students from the United States and 40 students from Turkey rated each of the 15 walks for mystery and surprise. Each group's ratings on each scale had high interobserver reliability. The students judged the level of mystery and surprise as intended: They rated the curved streets as higher in mystery than the straight ones, and when the difference between approach and post-turn streets increased, the judged surprise increased. For the evaluations, different respondents (34 from Turkey and 29 from the United States) walked through each of the 15 VEs assigned in random order. They rated each approach street and each post-turn street independently for interest and visual appeal. Because interest and visual appeal had high correlations, we combined them into a composite preference scale. For each country, preference increased with increases in mystery and surprise. To get at the dynamic experience, future research could employ continuous and nonreactive measures of response as people move through environments. (Contains 1 table and 15 figures.)