Previous research in spatial reorientation, which only presented the target location in the corner, has found that adults weighed angles more than wall lengths. We proposed that in previous research, angular cues were available for direct use whereas length cues had to be associated with the left/right sense. We thus investigated whether the directness of cues rather than the cues themselves accounted for the previous findings in the reorientation task. Through navigating a virtual environment, 111 participants were trained to remember target locations in a parallelogram-shaped room and tested in varied versions of the training environment: (a) a reverse-parallelogram environment where angular information conflicted with wall length information, (b) a rhombic environment that preserved only angular information from the training environment, and (c) a rectangular environment that preserved only wall length information. We varied the directness of the two cues in the current study. In addition to the condition with target location in the corner, we included a condition that placed the target positions in the middle of the walls, making the length cues direct. The results revealed that angular information no longer received priority, especially in the wall condition. More interestingly, compared to the group trained with the target positions at walls, the group trained with target positions at corners weighted angular cues more heavily in conflict trials and performed less accurately using length cues in rectangular environments. The results suggest that human adults prefer to rely on direct cues more than indirect cues in the reorientation process.