Most definitions of engineering give machines and mechanical objects a central role. Engineers are makers and users of mechanical objects in their environment. Research supports the notion that interactions with engineered artifacts enhance engineering learning. This study introduces a task simulating a real-world engineering application and uses this task to examine how aptitudes, interests, and direct manipulation of mechanical objects influence performance. We hypothesized that engineering students would generate better assembly instructions when they had the box of component parts (BOP) than when they had the engineering drawing only. We also hypothesized that student's mechanical aptitude (MA) and interests in things each would interact with experimental condition's impact on performance. First-year engineering students (N = 383) created assembly instructions in a mixed experimental and correlational design. A random half was assigned to create instructions with a drawing only, whereas the other half created with both a drawing and a box of component parts present. Assembly instructions were evaluated by professional engineers blind to experimental conditions. They rated instructions from the BOP group as superior to those coming from the control group. Students with greater mechanical aptitude received better evaluations, but there was no evidence the experimental variable was moderated either by mechanical aptitude or by thing orientation (TO). This study suggests that mechanical objects can enhance engineering instruction, especially when they are aligned with professional practice.