Paper Roller Coaster Challenge: A Low-Cost STEM Activity That Actually Works

One of the best things about teaching K-6 technology is that some of the most effective engineering challenges cost almost nothing. The paper roller coaster is a perfect example. Students build a functional roller coaster track out of paper strips, tape, and cardboard — and then test it with a marble.

That’s it. And it’s genuinely one of the activities my students ask to repeat.

Why It Works

The paper roller coaster isn’t just a craft project. Done well, it hits several real STEM concepts simultaneously:

Potential and kinetic energy. The height of the starting ramp determines how much potential energy the marble has. Too flat and it stalls. Too steep and it flies off the track. Students have to find the right balance through trial and error — which is exactly the kind of hands-on reasoning that sticks.

Force and motion. Friction, gravity, and momentum all come into play. A smooth curve keeps the marble on track; a sharp corner kills speed. Students discover this by building and testing, not by being told.

Engineering design cycle. Every group builds, tests, fails, and adjusts. The failure is the point. A marble that rolls off the track halfway through isn’t a bad result — it’s data.

Structural thinking. Paper doesn’t hold a curve well on its own. Students have to figure out how to fold, roll, and brace their strips to create channels stiff enough to support the marble. That kind of material problem-solving is core engineering thinking.

What You’ll Need

The materials list is short on purpose. Part of the challenge is building something functional out of very simple supplies.

• Paper strips (I cut standard printer paper into thirds lengthwise — each sheet gives you three strips)
• Masking tape or painter’s tape
• Cardboard or foam board for the base structure
• Marbles (one per group, plus extras for testing)
• Scissors
• A ruler for planning

Optional: cardboard tubes (toilet paper or paper towel rolls) for tunnels, index cards for platforms, small plastic cups for the catch basin at the end.

How I Run It in Class

Day 1 — Introduction and planning (30 minutes)

I start by asking students what they know about roller coasters. How do they start? Why do they go fast at the bottom? What would happen if the track had a sharp corner instead of a curve?

Then I show them the video below, which demonstrates what a finished paper roller coaster looks like and gives them a sense of what’s possible with just paper and tape. We watch it once without pausing, then I pause it at key moments to ask questions.

Watch the video

After the video, groups sketch a rough design on paper. They have to plan at least: the starting height, one curve, and a landing zone. I don’t enforce a specific layout — the point is that they have to think before they build.

Day 2 — Build and test (45–60 minutes)

Groups build. I circulate and ask questions rather than giving answers: “Why do you think the marble keeps falling off there?” “What would happen if you made that ramp longer?” “How could you hold that curve open without using so much tape?”

First tests almost always fail in interesting ways. I treat every failure as a share-out moment. “Group 4’s marble made it all the way to the loop and then stopped — what do you think caused that?” Other groups learn from watching, and the group that failed figures out their fix faster because they’ve had to explain the problem out loud.

Day 3 — Refine and demonstrate (30 minutes)

Groups make their final adjustments and then run an official demo. The marble has to complete the full track without being guided by hand. Groups that succeed can extend their track or add a feature (tunnel, loop attempt, double ramp).

What I’d Do Differently

The biggest mistake I made the first time I ran this: I let groups use too much tape. When tape is unlimited, students tape everything to a flat table and the design becomes rigid too fast. Now I give each group a fixed tape budget (usually a 12-inch strip to start, with more available if they can justify why they need it). The constraint forces better engineering decisions.

I’d also recommend building your own prototype before class. You’ll find the tricky parts — where the paper wants to collapse, how long the strips need to be for a smooth curve — and you’ll be able to ask better questions when you see students hitting the same problems.

Standards Connections

This activity maps naturally to Next Generation Science Standards (NGSS):

• 3-PS2-1 and 3-PS2-2 — Planning and conducting investigations about the effects of balanced and unbalanced forces
• 4-PS3-1 — Using evidence to construct an explanation of the relationship between the speed of an object and the energy of that object
• K-2-ETS1-1 through K-2-ETS1-3 — Engineering design: defining problems, developing solutions, comparing solutions

For older students (grades 4–6), add a reflection component where groups calculate the height of their starting ramp and predict whether it provides enough energy to complete a loop of a given radius. Even rough estimates lead to good conversations about the relationship between height and speed.

Extensions

If groups finish early or want to go further:

• Loop challenge: Can you add a vertical loop? (Harder than it looks — the entry speed has to be high enough and the loop has to be smooth.)
• Speed test: Time how long the marble takes to travel the full track. Then modify the design to make it faster. What changes made the difference?
• Multi-marble relay: Two tracks side by side, timed race. Which design wins and why?
• Structural analysis: Before dismantling their finished track, groups sketch exactly how they built each curve and joint, annotating what worked and what they’d change if they built it again.

The Real Value

I keep coming back to this challenge because it’s one of the few activities where the frustrating parts — the marble falling, the track collapsing, the curve being too tight — are also the most educational parts. Students don’t need me to tell them what went wrong. They can see it. They want to fix it. That self-correcting loop is what engineering design feels like in practice.

And it costs about $2 in materials per class.