Science Kits for Kids

Consumable, semi-consumable, and permanent

Everything in this category falls into one of three buckets, and the bucket matters far more than the brand.

Consumable kits are used up. A geode kit publishes ten geodes; once cracked, they are cracked. A crystal-growing kit grows one crystal. These are experiences, not toys, and they are excellent value judged as experiences and poor value judged as toys. The failure mode is entirely in the framing: a child told they have a science kit expects it to still be a science kit in February.

Semi-consumable kits use up chemicals but keep the equipment. Thames & Kosmos’s C-series is the clearest example: the C500 publishes 28 experiments, a 48-page manual and five chemicals, and the company sells replacement chemicals separately. Bought with that in mind, it is a good deal — the glassware, the spatula and the safety glasses carry into the next set up the line. Bought as a permanent toy, it disappoints in a month.

Permanent equipment does not run out at all. A microscope is the standout: the specimens are free and infinite, because they are the garden. A telescope is the same, more so — the sky changes every month whether or not anyone opens the box. These are the only purchases in this category that pass the Replay Signal outright, and it is why they sit at the top of our picks despite costing more up front than most kits.

Reading a science kit’s published numbers properly

Experiment counts are the headline number and they need care. A kit advertising “100 experiments” and one advertising “28 experiments” may deliver similar amounts of actual science, because there is no standard for what counts as one. Look instead for two published figures that are much harder to inflate: the manual page count, and the number of distinct chemicals or components.

Manual pages are a decent proxy for depth. A 48-page manual covering 28 experiments is giving roughly a page and a half to each, which is enough for a method, an explanation and a why. A kit claiming a hundred experiments in a twelve-page leaflet is claiming something else.

Component counts are the other honest number, and they matter because real science kits work by recombination. Five chemicals is not five experiments, it is a small combinatorial space, and a good manual walks through it. That is exactly what the C-series does, and it is why Thames & Kosmos publishing its complete manuals free on its own site is unusually useful — you can read the entire experiment list before you spend anything.

Safety, said properly rather than nervously

Chemistry sets sold in the US and Europe are made to published standards. Thames & Kosmos states compliance with EN 71-4, the European standard specifically for chemistry experiment kits, and the general US standard for toy safety is ASTM F963. Those are real documents with real requirements, and a kit that names the standard it meets is telling you something a kit that says “safe for kids!” is not.

What the standards do not do is tell you whether your child should be doing it alone. That is a judgment, and our position is boring: at the bottom of a 10+ rating, sit with them. Not because the chemicals are dangerous when used as directed, but because the first time a reaction does something unexpected is the moment the science actually lands, and it lands better with someone to say it to.

Where subscriptions fit

A subscription solves the consumable problem by leaning into it: each crate is used up, and another arrives. KiwiCo publishes a clear ladder by age — Kiwi Crate at 6 to 9, Tinker Crate at 9 to 12, Eureka at 12 and up — and MEL Science publishes two to three experiments per monthly set for ages 10 and up.

The honest framing is that you are buying twelve projects, not one toy, and it should be priced against twelve months of anything else rather than against a box on a shelf. Where subscriptions genuinely win is for a child who likes the ritual as much as the science. Where they lose is when nobody is around to do the project on the day it arrives, and three unopened boxes accumulate on top of the fridge.

For a first purchase, we usually route people to a microscope instead — permanent, open-ended and cheaper across a year than most subscriptions. If the child is already interested and you want to keep feeding it, the subscription page does the annual arithmetic properly.