A spirit level sounds like a builder’s tool, and mostly it is. But on a survey it earns its place as a measuring instrument. When a client says a door has started catching, or a marble rolls across the lounge, or the extension “feels like it’s going downhill”, the level is what turns a feeling into a figure. One door frame 10mm out of plumb is a note in a report. Every frame on the same side of a Victorian terrace leaning the same way is a chapter — and possibly a conversation about movement.
The same instrument checks whether a kitchen worktop was set level or the units are heading for the corner, whether a damp-proof course line actually holds its 150mm above ground the whole way round, whether a bathroom floor has any fall towards its drain, and which way a suspended timber floor is dipping when a sleeper wall has settled. None of that needs laser kit. It needs a straight, rigid length of aluminium with a vial you can trust.
And trust is the whole game. The difference between a £12 level and a £90 one is not that one shows a bubble and the other doesn’t — it’s whether the bubble is telling the truth, whether it still will be after a year in the van, and whether you can prove it in thirty seconds. That last part matters most, and it’s where this guide starts and ends.
The one I’d buy
Stabila 196-2 (120cm)
The box level the trades standardised on, for the same reason the damp industry standardised on the Protimeter: it stays accurate. A rigid box section, a vial rated at 0.5mm/m and famously repeatable, and removable die-cast end caps so you can work right into corners. Buy the 120cm as the workhorse; the range runs from 40cm to well over 2m if it ever needs company.
Typically £75–£100 for the 120cm
Check price on AmazonHow I judge a spirit level
A published accuracy figure, in mm/m. Decent levels state their vial accuracy as millimetres of error per metre of length — 0.5mm/m is the professional standard, and Stabila, Milwaukee and the better trade brands all publish it. Across a 1.2m level that’s 0.6mm of permissible error: less than the thickness of a pound coin, and far tighter than the construction tolerances of any house I survey. A level that doesn’t publish an accuracy figure at all is telling you something. And treat “0.029°” on a spec sheet as the same claim in different clothes — it’s 0.5mm/m converted to degrees.
Length — more than you think. A level reads the two points its ends happen to sit on, so a short level on a lumpy surface reports the lump, not the structure. For anything structural — door frames, walls, lintels — 600mm is the sensible minimum. For floors, worktops and DPC lines you want 1200mm, and 1800mm if falls across whole rooms are the question. A torpedo level has its uses in a survey bag (pipe runs, gully falls, appliance feet), but it is not the tool for judging a building.
It must pass the flip test. Any level can be checked for calibration in thirty seconds without any reference surface: set it down, note precisely where the bubble sits against the lines, then turn it end-for-end onto exactly the same spot. A true level shows the identical reading both ways; any difference is twice the vial’s error. A level that fails is not a level — it’s a straightedge with delusions.
Readability in bad light. Survey work happens in unlit hallways, cellars and lofts, reading vials at skirting level and above head height. Vial magnification, high-contrast ring markings and a bubble you can read from above rather than straight-on all sound like brochure fluff until you’re squinting at a worktop in a north-facing kitchen in December.
Rigidity — because a level is also a straightedge. Half the diagnostic value of a 1200mm level is the gap under it. Bridge a floor and the daylight beneath tells you hump or dip, and roughly how much once a tape is against it. That only works if the level itself is straight and stays straight, which is why box-section profiles beat I-beams and why the reinforced-spine designs exist. A level that’s spent a year diagonal under a van shelf needs the flip test more often.
The comparison
| Model | Stated accuracy | Format | Typical price | Best for | |
|---|---|---|---|---|---|
| Stabila 196-2 (120cm) | 0.5mm/m | Box section, removable end caps | ~£85 | The all-round professional benchmark | View |
| Milwaukee Redstick Backbone | 0.5mm/m | Reinforced-core box, magnetic option | ~£90–£140 | Hard site use; steelwork with the magnets | View |
| Stabila TECH 196 DL | Vial 0.5mm/m; digital ±0.05° at 0°/90° | Box section, two digital displays, IP67 | ~£180–£300 | Recording falls and slopes in reports | View |
| Stanley FatMax Pro (1200mm) | 0.5mm/m | Box section, shock-absorbing end caps | ~£40–£65 | Mainstream trade quality, easy to buy | View |
| OX Pro Level (1200mm) | 0.5mm/m | Box section, rubber-capped ends | ~£40–£55 | Strong value from a UK trade brand | View |
| Stabila 70-2 (120cm) | 0.5mm/m | Slim rectangular section | ~£25–£40 | Best on a budget — the honest workhorse | View |
| Huepar 600mm digital | Digital ±0.05° at 0°/90° (claimed) | Box section, LCD, magnetic base | ~£40–£55 | Budget digital — degrees and percent on a screen | View |
The picks in detail
Stabila 196-2 (120cm)
Best overall — the benchmark, and my top pick above
Ask a joiner, a bricklayer and a shopfitter to name a level and you’ll hear the same word three times. The 196-2 is a stiff aluminium box section with an electrostatic powder coat that wipes clean of plaster and mortar, a horizontal vial Stabila rate at 0.5mm/m in any position, and end caps that come off so the level works right into a corner — useful against door linings and skirtings. Owner feedback is remarkably consistent on the thing that matters most: these hold their calibration through years of site use, which is exactly what you want from an instrument that lives in a van.
The 120cm length is the surveyor’s sweet spot — long enough for floors, worktops and DPC lines, short enough to travel behind the seats. It isn’t cheap for what looks like a yellow stick, but the price buys the one property a level exists to provide: a reading you don’t have to second-guess.
For
- 0.5mm/m vials with a long-standing reputation for holding calibration
- Rigid box profile doubles as a trustworthy straightedge
- Removable end caps for working into corners
- Huge range of lengths if you standardise on the system
Against
- Serious money next to the DIY shelf
- No magnets or digits at this price — those are separate models
Typically £75–£100
Check price on AmazonMilwaukee Redstick Backbone (120cm)
Best for hard use — and the magnetic edition for steelwork
Milwaukee’s answer to the Stabila, built around a reinforced metal core — the “backbone” — that stiffens the box against exactly the abuse that puts levels out of true: twisting, dropping, and years under other tools. Accuracy is the same professional 0.5mm/m, and the SHARPSITE vial is genuinely easier to read at an angle or in poor light, which owner feedback singles out repeatedly. The magnetic edition adds rare-earth magnets along the milled face — niche for house surveying, quietly brilliant the day you’re checking a steel beam, a lintel or a scaffold handrail and need both hands for the camera and the notebook.
For
- Reinforced core built to stay straight and stay calibrated
- High-visibility vial that reads well in dim interiors
- Magnetic version grips steel while you write or photograph
Against
- All that metal makes it noticeably heavier than the Stabila
- Magnetic editions carry a meaningful premium
Typically £90–£140 depending on length and magnets
Check price on AmazonStabila TECH 196 DL digital
The professional digital pick — when the report needs a number, not a bubble
A 196 box level with two illuminated digital displays set into it, reading out in degrees, percent or mm/m. That last pair is the surveyor’s case for it: a bubble tells you a floor is off, but this tells you the floor falls 1.3° — or 11mm per metre — towards the bay, and that number goes straight into the report with a photograph of the display as evidence. Stabila quote ±0.05° for the electronics at level and plumb, the conventional vials are still there when the batteries die, and the IP67 rating means rain and brick dust aren’t the end of it. Available in 61, 122 and 183cm; the 61cm makes a sensible first digital level, the 122cm is the one for floors.
For
- Records actual falls in degrees, percent or mm/m — report-ready
- Illuminated displays readable in cellars and lofts
- Proper vials still fitted, so it never stops being a level
- IP67 dust and water protection
Against
- The most expensive thing on this page by a distance
- Overkill if you never need to write the slope down
Typically £180–£300 depending on length
Check price on AmazonStanley FatMax Pro box beam (1200mm)
The mainstream trade choice — proper spec, buyable everywhere
Stanley’s top-tier box beam publishes the same 0.5mm/m accuracy as the premium German kit at roughly half the money. The spec sheet reads sensibly for site life: rigid box profile, three vials on the 1200mm, and shock-absorbing end caps for the inevitable drops. Published user feedback is solid on straightness and readability; the recurring caveat, fairly, is that it’s a heavier lump than its price peers. If the Stabila feels like an indulgence and the no-name shelf feels like a gamble, this is the grown-up middle.
For
- 0.5mm/m claimed accuracy at a mid-range price
- Shock-absorbing end caps — levels get dropped
- Stocked everywhere, so replacement is painless
Against
- Heavier than most rivals at this length
- Owner feedback on long-term calibration is good but less storied than Stabila’s
Typically £40–£65
Check price on AmazonOX Pro Level (1200mm)
Best value trade level — the bricklayers’ favourite underdog
OX is a UK trade brand that made its name in brickwork tools, and its Pro levels have built a strong following on merit: a box section with 0.5mm/m-rated acrylic vials, decent end protection, and a price that undercuts the big names by enough to notice. Owner feedback from trades is consistently positive on accuracy out of the box and on the clear, slightly oversized vials. It doesn’t have Stabila’s decades of calibration folklore, which is really the only thing separating them — run the flip test at purchase and periodically after, as you should with any level, and it’s a lot of instrument for the money.
For
- Professional spec at a value price
- Clear, high-visibility vials
- Well-regarded UK trade brand with a full range of lengths and sets
Against
- Less long-term track record than the premium German brands
- The blue paint is a matter of taste
Typically £40–£55
Check price on AmazonStabila 70-2 (120cm)
Best on a budget — Stabila vials without the Stabila price
The 70-2 is the quiet bargain of the range: the same 0.5mm/m vial technology as its dearer siblings in a slimmer, lighter rectangular section. What you give up against the 196-2 is rigidity and the removable end caps; what you keep is the part that makes a level a level — a vial you can trust, from the company with the best reputation in the business for them, at a price that regularly dips under £30 for the 120cm. For a homeowner, a landlord, or anyone furnishing a first survey bag, this is the answer.
For
- Genuine Stabila 0.5mm/m vials at a budget price
- Light enough to carry all day without noticing
- The cheapest route to a trustworthy 1200mm reading
Against
- Slimmer profile flexes more than a box section
- No removable end caps; fewer premium touches
Typically £25–£40
Check price on AmazonHuepar 600mm digital level
Budget digital — degrees and percent on a screen for sensible money
Huepar made its name in budget laser levels, and this 600mm digital level is the value route to what the Stabila TECH does: an LCD reading in degrees, percent or mm/m, conventional vials alongside, a magnetic base and an audible beep at dead level. The claimed accuracy for the electronics is ±0.05° at 0° and 90°, softening in between — adequate for describing a floor fall or a leaning frame. Owner feedback is good for the money, with the caveats you’d expect of the class: check the zero against the flip test on arrival, and don’t expect Stabila build. At 600mm it’s short for floors — pair it with a straightedge, or use it where it shines: falls, ramps and short runs.
For
- Digital degrees/percent readout at a fraction of flagship money
- Magnetic base and audible level alert
- Vials still work with the screen off
Against
- 600mm is short for judging whole floors
- Budget-class consistency — verify the calibration on arrival
Typically £40–£55
Check price on AmazonWhat actually matters — and what doesn’t
The flip test is the only spec you can verify yourself. Every claim on the packaging reduces to one question you can answer in the shop: does it read the same both ways round? Do it before you buy, again after any serious knock, and to the old level in the shed before trusting it with anything that matters. Vials drift and adhesives age, and a level that has quietly gone out of true is worse than no level — it hands you confident wrong answers.
Digital earns its keep when the number must be written down. For a builder, level is a yes/no question and a bubble answers it. For a surveyor, the magnitude is the finding. “The ground floor falls approximately 20mm per metre towards the rear left corner” is a sentence that supports a conclusion about settlement; “the floor is a bit off” is not. A digital level gives you that figure directly, in mm/m or percent, and the display photographs well as file evidence. If your work never requires the number, save the money and buy the better vial.
Falls are smaller than you think. A wet-room floor laid to a 1-in-80 fall — a common design figure — slopes at 1.25%, or about 0.7°. You cannot judge that by eye, and a standard vial only barely declares it. This is exactly the territory where a digital readout stops an argument between a client and their tiler, in either direction.
What the display stand pushes that you can skip. Vial counts, mostly — a third or fourth vial at 45° is decoration for house diagnostics. Rubber “comfort” grips, painted-on angle scales, and combined laser-and-vial gadgets that do neither job well. Magnets are genuinely useful if you regularly meet steel and an irrelevance if you don’t — and they collect every screw and filing in the van. Spend the marginal pound on the vial, the box section and the brand’s calibration reputation, in that order. One honest caveat the other way: a digital level with a user-zero function can be zeroed badly as easily as well, so learn its calibration routine and check it against the flip test rather than trusting the button.
Questions I get asked
How accurate is a spirit level? A professional-grade level is rated at 0.5mm/m — over a 1.2m level the bubble is permitted to be wrong by 0.6mm, far tighter than any tolerance that matters in an existing house. Budget levels either meet the same figure (the Stabila 70-2 does), quote a looser one, or say nothing at all — which usually answers the question. In practice the spec matters less than the condition: an expensive level that has been dropped can read worse than a cheap one that hasn’t, which is why the flip test beats the brochure.
How do I check if my spirit level is accurate? The flip test, and it costs nothing: put the level on a firm surface that’s close to level, note exactly where the bubble sits against the graduation lines, then rotate the level 180° end-for-end and set it back on the same spot. The two readings should be identical; if they differ, the vial is out by half the difference. Check the plumb vial the same way against a door frame, flipping face-for-face. Digital levels usually include a two-position calibration routine that automates the same idea — run it on a solid surface, not a wobbly table.
What length spirit level do I need? One length rarely covers everything, but if forced to choose I’d take 1200mm: long enough to make honest judgements about floors, worktops and DPC lines, short enough to travel. Go no shorter than 600mm for anything structural — door frames, walls, lintels — because a short level reads local bumps rather than the element. For whole-floor falls, 1800mm or a 1200mm level on a straightedge is better still. Torpedo levels (around 230mm) are for pipework, appliance feet and gully falls, not buildings.
Are digital spirit levels worth it? If you only ever need to know whether something is level, no — a good vial answers that faster and never needs batteries. They earn their price when you need the actual value: recording a floor’s fall in a survey report, setting or checking a designed gradient (a 1-in-80 shower fall, a ramp, a drainage run), or photographing evidence of a lean. The professional option is the Stabila TECH 196 DL; the Huepar covers the same jobs at a budget, with more diligence required on calibration. Either way the vial levels stay in the bag — digital supplements a good level, it doesn’t replace one.
If the reason you’re reading this is a door that’s started sticking or a floor with a lean you can feel through your socks, measuring it is the easy part — saying what it means is the day job. The practice does building surveys across Sussex, Kent and Surrey where distortion, settlement and movement are assessed in the context of the whole building — get in touch. And if you’re furnishing a survey bag, there are more field-kit guides, including one on damp meters — the other instrument every odd floor seems to end up needing.
Small print. This guide is general information, not a survey of your property; an out-of-level floor or out-of-plumb frame has many possible causes, and a level reading alone is never a diagnosis. Prices are indicative, move constantly and are checked at the time of writing; the linked page always shows the current price. As an Amazon Associate I earn from qualifying purchases.