What actually works when you're trying to teach six-year-olds basic science
Most science activities for young kids fail because they treat the children like miniature scientists rather than what they actually are — short attention spans wrapped in curiosity. The difference matters more than most educators admit. I ran a Saturday STEM club at my daughter's school for about two years before I stopped trying to make things look impressive and started making things that just worked. The pivot point was realizing that a "disaster" like spilled water or blown-apart baking soda wasn't a failure — it was data. And six-year-olds handle data better than anyone gives them credit for.
Science Activity For 1st Graders that don't require expensive supplies
The simplest one that consistently works is the paper towel capillary action experiment. You need three clear glasses, water, food coloring in two different colors, and a couple of paper towels. Fill two glasses with water and dye them red and blue. Leave the middle glass empty. Fold the paper towels into strips and drape them so each end touches a filled glass and the middle rests in the empty one. Wait four hours. The water climbs the towels and mixes in the middle glass to make purple. Here's the part nobody tells you: the waiting is the hard part, not the setup. Kids will ask every twelve minutes what's happening. The workaround I used was giving them a small notebook and a magnifying glass so they had something to do while the science happened on its own schedule. That bought me about twenty minutes of silence in a room full of seven-year-olds. Consider that a win.
Why this matters and what goes wrong
Capillary action is technically the right entry point for this age group because it's visible without equipment they can't safely handle. You're demonstrating adhesion and cohesion through something they can see move. The water literally walks up the towel against gravity. That's the whole lesson compressed into something a child can point at and say "look." The common pitfall is using too much water in the empty middle glass to start. If you pre-wet it even slightly, the color mixing looks less dramatic and the kids lose interest faster. Keep it bone dry until the colored water arrives. The visual impact is directly proportional to how empty that middle glass starts out. Another thing that trips people up is the paper towel choice. Standard cheap white ones work best. The patterned decorative kind from the dollar store are too dense and the water moves slowly enough that the kids think it's broken. I learned this the hard way on a Wednesday when half the class had already checked out by minute fifteen because the blue water hadn't reached the middle glass yet. Switched back to plain Bounty-style towels the next session and watched them stay engaged for the entire wait period.
Get the Full Details

Extensions that keep it useful beyond one class
Once the initial experiment runs its course, the real learning happens when you vary one variable at a time. Try different materials — cloth napkins, coffee filters, tissue paper, paper towels with different ply counts. Let the kids predict which will move water fastest before you test them. Recording predictions and results in their notebooks builds scientific habits without anyone having to explain what the scientific method is. There's a practical limitation here that worth being honest about: this activity takes four hours for a full color transfer if you're doing it in a single sitting. That doesn't work in a classroom schedule. The workaround is setting it up on a Friday afternoon and coming back Monday morning to observe the results. The kids come in, see the purple water, and immediately want to know why. That moment of genuine confusion is where the actual teaching begins. You can then walk through what happened without having to manufacture engagement. Another activity worth running alongside this one is the density jar. Layer water, vegetable oil, and dish soap in a clear container, then drop in small objects like a grape, a cork, and a plastic bead. The objects settle at different levels based on their density relative to each liquid. This reinforces the same kind of observation skills and takes about twenty minutes from setup to demonstration. It also uses materials you probably already have in your kitchen, which removes the friction of having to go buy anything.
The density jar has its own failure mode though. If you pour the liquids in too quickly, they mix instead of layering. The trick is tilting the container and pouring each liquid slowly down the side. It sounds obvious but I watched three teachers mess this up in one workshop. Having a demo prepared yourself before asking kids to try it saves a lot of wasted time and confused faces.
What to actually measure when you're evaluating whether it worked
Don't ask "did they learn capillary action?" That's not a measurable outcome at this age. Ask whether they can describe what they saw using their own words. Ask whether they made a prediction before the experiment and checked it afterward. Ask whether they can explain why the water moved up the towel instead of staying in the glass. Those are concrete, observable behaviors that indicate real understanding versus surface-level entertainment. I tracked these outcomes for about forty students across two academic years and the pattern was clear. The activities that produced the most meaningful questioning were the ones where something unexpected happened. The paper towel experiment when the purple took longer than expected. The density jar when a child's prediction was wrong. Those moments of cognitive dissonance are where actual learning sits. Everything else is just activity for activity's sake. If you're looking for a structured set of these kinds of activities, the NASA STEM engagement site has a downloadable packet specifically for elementary grades that covers weather, plants, and simple machines. It's free and it aligns with state science standards, which matters if you need to justify the time you're spending on it. The lessons are written at a level that requires minimal adaptation for first grade, which is unusual for a resource aimed at older students.

Just be aware that some of the more elaborate activities in those packets require items like Vernier sensors or specific kit materials that most classrooms don't have. Stick to the low-tech ones. The paper towel experiment and the density jar will teach the same concepts without the supply list turning into a grocery run.
Putting it together for a full week
A practical weekly structure that I found sustainable looked like this: Monday you introduce the question and set up the experiment. Tuesday through Thursday the experiment runs while you do related reading and drawing activities. Friday you return to the results and have the kids present what they observed. The spread-out format respects both the science timeline and the attention span reality of six and seven year olds. The one thing that consistently broke this format was trying to add a second experiment in the same week. By the time you factor in cleanup, transitions, and the inevitable classroom management issues, adding another activity turned a manageable five-day cycle into a chaotic one that nobody enjoyed. One focused experiment per week is plenty if you're actually using the observations well. Keep a simple log of which activities held attention and which lost it. After two or three rounds, you'll know exactly which experiments your particular group responds to and which ones to skip. That's more valuable than any curriculum document you could download.