Shoes, socks or barefoot on a walking pad?

This question gets answered confidently all over the internet and almost never with a citation, so here is the whole of what we could read from the makers of the machines we track, on 2026-10-13.

One instruction exists. A KingSmith WalkingPad manual (model ORWKPG1) states, on its English safety page: “Please wear sports shoes and clothes.”The French edition of the same manual carries the identical line. That is it — no sole type, no tread requirement, no explanation, and no statement about bare feet either way.

The brand’s public FAQ does not mention footwear at all.We read walkingpad.com’s FAQ in full looking for shoes, socks, bare feet or slipping, and the document contains none of them. This is the same FAQ that will tell you about continuous-use limits and power cords, so the absence is not for lack of space.

One maker does not publish its manual as text.Sunny Health & Fitness’s manual page for the TreadPad model we track offers the document by email or as an online preview rather than as readable page content, so there is no footwear line there for us to quote. We are recording that as unread rather than as absent, because those are different things.

That is a thin evidence base for a question people ask constantly, and it is why the rest of this page argues from the machines’ own published geometry rather than from a recommendation nobody has published.

Socks are the option we would actively argue against, and the argument is built from published figures rather than from caution.

There is nothing to hold. Exactly one of the 8pads we track publishes a handle — the UREVO Strol 2E — which means that on every other machine here, a slip has no recovery hardware at all. On a gym treadmill you grab a rail. On a walking pad you step off a moving belt. The pads with a handle, which is a one-item list.

The belt is narrow. Published belt widths across the pads we track run 15 to 16.5 inches, and belt lengths 27.6 to 47.6 inches. A sock foot drifting an inch sideways on a 15 inch belt is most of the way to the edge.

Those two findings pull against each other, and it is worth saying so rather than tidying it away: the one pad here that publishes a handle is also the one with the narrowest published belt, at 15 inches. There is no machine in this registry that gives you both something to hold and the widest walking surface.

And there is a documented way the deck becomes slippery. KingSmith’s own lubrication article warns that over-lubricating pushes surplus silicone up to the top surface of the belt. That is the one failure in this category that genuinely makes the walking surface slippery, it is caused by maintenance rather than by wear, and it is far worse in socks than in shoes. How much silicone the maker actually asks for — it is one to two drops.

Smooth sole, narrow moving belt, nothing to grab, and a documented route to a slick surface. Of the three options in the title, this is the one with no argument in its favor beyond comfort.

Barefoot has a real case, and it is worth stating properly rather than dismissing: bare skin grips better than a sock, you can feel where you are on the belt, and at the speeds a desk pad actually gets used at — the slow end of a range that tops out between 3 and 6.2 mph across the machines we track — the consequences of a stumble are small.

What it costs: no protection at the front roller, where a moving belt meets a fixed housing; no structure under a foot that is doing an hour of repetitive work; and a belt that is designed as a drive surface rather than a floor. It is also the option furthest from the only maker instruction that exists, which asks for sports shoes.

If you are going to do it anyway — and plenty of people walking at 1 mph beside a desk will — the sensible version is the bottom of the speed range, a machine with something to hold if you have one, and a clean belt. Which leads to the one genuinely useful thing the makers do publish about feet.

The instruction says “sports shoes” and stops. We are not going to turn four words into a specification, so here is the honest split between what is published and what is inference.

Published: KingSmith’s maintenance article names rubber from shoes, along with airborne dust and hair, as what reaches the belt and the motor, and asks for a weekly wipe-down of the belt and frame plus vacuuming around the machine. That is a direct argument for one clean indoor-only pair rather than whatever you wore outside, and it is the maker’s argument rather than ours. The maintenance jobs in order of what they actually prevent.

Inference, clearly labeled as ours: on a belt 15 to 16.5 inches wide, a stable sole that does not overhang the belt edge matters more than cushioning, because the failure mode here is lateral rather than vertical. A shoe you can keep on for an hour beats a shoe that is better in every specification and comes off after twenty minutes.

What we will not tell you is a heel-to-toe drop, a sole compound or a model. No maker in this category publishes a footwear specification, we have not walked on these machines, and a shoe recommendation dressed up as a spec is exactly the kind of invented authority this site exists to avoid.

Here is the part of the footwear question that is actually a measurement problem, and it is the reason this page sits next to the desk pages rather than the safety ones.

A walking pad raises you by its in-use height before you put anything on your feet. A shoe sole raises you again. If the machine lives under a desk, the number that decides whether the setup works is the gap left above your foot at the top of a step — not the gap above the bare deck, which is what every listing implicitly quotes.

So measure it with the shoes on. Stand on the deck in the pair you will actually wear, and measure from the top of your foot to the underside of the desk. If that number is uncomfortable, the fixes in order of cost are: a thinner-soled indoor pair, a taller desk, or a pad with a lower deck. The clearance arithmetic for every machine we track and the pads ranked on deck height for a standing desk are the two pages that do this with published numbers.

We have not walked on these machines in shoes, in socks or in bare feet, and we have not tested a belt surface for grip. Every maker statement above is quoted from the document named in the sources with the date we read it; where a document could not be read, it is recorded as unread rather than as silent. Every machine figure is the maker’s own, computed from our registry at build time. The lateral-stability argument is clearly marked as our inference, because that is what it is. The full method.