If you're a small lab or a solo engineer and you type “which Fluke multimeter should I buy” into a search box, I'm going to say something that might annoy you: you're asking the wrong question.
Not because Fluke makes bad instruments. Quite the opposite. Because the question already assumes the answer is a Fluke before you've looked at what you're measuring. And in my line of work, I review test equipment before it reaches customers — roughly 200 units a year — and I've watched too many smart engineers pick a perfectly good brand that was the wrong instrument for the actual measurement.
It took me about four years and 150+ specification reviews to understand that fully. At first I trusted the famous names. Then I started reading the “Specifications” page the way a quality inspector is supposed to, and I slowly stopped caring about the logo on the bezel. Brand is a shortcut, not a spec.
Start With the Class of Measurement
Let's walk through a concrete example. Say you're testing battery protection circuits, and you need to measure the voltage drop across a 1 mΩ shunt resistor at 1 mA. That's 1 µV. A 4.5-digit handheld meter — whatever brand, including the good ones — can't resolve that signal on its lowest range. You need a benchtop instrument with a 100 mV range and enough low-level resolution to see the change.
A Keithley 2010 multimeter, for instance, is a 7.5-digit bench DMM with a published basic DC voltage accuracy of ±0.002% of reading and 10 nV resolution on its most sensitive range. A Fluke 87V is an excellent 4.5-digit handheld with a CAT III 600 V rating, and it still can't see a 1 µV signal because it isn't designed to. That's not a knock on Fluke; it's physics. They belong to different measurement classes.
And if the test is impedance instead of DC voltage — say you're measuring capacitors or inductors at a specific test frequency — even a 7.5-digit DMM is the wrong class. That's when a Keithley LCR meter (or any LCR meter, really) is the right choice, because its accuracy is specified at defined AC frequencies and drive levels. The principle stays the same: the measurement class decides the instrument class.
So if your work lives on a bench and involves sensors, shunts, and voltage references, the honest answer to “which Fluke should I buy” might be: none of them. You want a bench DMM — maybe a Keithley 2000 or 2010, or a newer DMM6500, depending on budget. And there's nothing wrong with that: knowing which class you need is 80% of the buying decision.
The Small-Lab Excuse Is a False Economy
The pushback I hear next is completely fair: “I don't run a big company. I can't justify a few thousand dollars on a benchtop meter.”
And I have real sympathy for that. In 2022, I bought an $89 no-name meter to “make the budget stretch.” It showed 5.083 V when my 5.000 V reference was sitting right next to it. I sent it back, borrowed a six-and-a-half-digit unit, and spent a solid three weeks doubting every measurement I'd recorded up to that point. Honestly? So glad I learned the lesson on an $89 loss instead of a $22,000 redo — but it shook my confidence more than I'd like to admit.
Look, small doesn't mean cheap. It means right-sized. A used Keithley 2110 — a 5.5-digit benchtop multimeter — is a solid buy from any reseller who supplies a calibration certificate with NIST traceability. Used 2000-series meters appear on the market regularly as labs upgrade. Sometimes the only thing “wrong” with them is a dusty case.
And here's the part I actually care about as a quality person: spending $200 on the right instrument beats $89 on the wrong one every time. The small order is not a small measurement. The vendor who treated my first $200 order seriously is the one I sent a $20,000 order to a year later. I've seen that pattern repeat enough times that it's not a coincidence — it's the closest thing to a customer-loyalty spec I've ever found.
“The vendor who treats a $200 order seriously earns the $20,000 order. I've seen it happen more times than I can count.”
That's the customer-friendly side of a quality system. A $200 order might not pay anyone's bills, but the engineer who buys it today is the person specifying a new lab next year. I've learned not to guess which orders will grow; it's cheaper to treat all of them like they matter.
The Spec Sheet Is a Contract
Here's the shift that changed how I buy: a published specification is a promise. Turn to the accuracy table on any Keithley datasheet and you'll find temperature ranges, warm-up times, calibration intervals, and traceability statements. That's a contract. An $89 meter's “±0.5%” with no conditions, no temperature curve, and no traceability note? That's not a spec, it's a hope (and, honestly, often a decimal point away from reality).
I see the same pattern beyond multimeters all the time. When someone asks me about handheld thermal cameras, I ask what NETD and spatial resolution they need for the fault they're hunting. When someone asks about CMM arms, I ask which max permissible error matches their part tolerances. The measurement leads; the product follows. A camera with a famous name can't see a half-degree difference if its noise floor is 100 mK, and a $30,000 measuring arm won't hold your ±0.025 mm tolerance if its spec says ±0.050 mm. (We check dimensional conformity too, by the way. Surprise, surprise — dimensional specs can be as aspirational as electrical ones.)
There's a reason “keithley” shows up alongside so many low-level measurement questions: Keithley publishes the measuring conditions that other datasheets gloss over. You can verify a 2010's nanovolt-level resolution, input protection, and offset current — and you can hold them to it when the unit arrives. That's what a spec sheet is for. Trust is earned on the datasheet, not on the brochure.
One of my favorite things this year is explaining to a customer that a CAT rating is not an accuracy rating. He argued that his CAT IV clamp meter was “more accurate” because it was rated for higher voltages. I put the two datasheets side by side: the benchtop DMM beat his clamp meter on basic DC accuracy by well over a factor of 100. Safety certification and measurement precision are two different columns on the datasheet. People blur them all the time — and vendors don't always rush to correct them. It's one of the first things I check now when a customer brings me a meter they're proud of.
Yes, You Should Still Own a Handheld
Let me answer the objections I can hear forming. First: “I work on a plant floor, not a bench.” Fine. That's a different measurement class, and a rugged handheld with CAT III/IV ratings is the right tool. Keep it. But don't expect it to tell you anything useful at microvolt levels, and don't cite its safety rating in a discussion about accuracy. Those are two different specs pretending to be one.
Second: “I can't afford benchtop precision.” I've already shared my answer: the used market and entry-level models exist for a reason. Buying the right used meter with a calibration certificate beats buying a new meter with the wrong resolution, every single time. There's a reason the used 2000-series has a following — and it's not nostalgia. Those units hold calibration, and Keithley still supports them. My experience is based on about 200 bench and portable units a year, mostly electrical measurements. If you work in high-voltage power, your priorities will lean toward isolation and safety — the decision order, however, stays the same. Also worth saying: I don't do calibration in-house, and I won't pretend to quote official uncertainty budgets. That's why we send instruments out to an accredited lab annually. Knowing where your competence ends is part of measurement quality, too.
Let the Measurement Pick the Instrument
I was one return shipping label away from a decade of quietly bad data. That's not an exaggeration. The $89 meter taught me that looking like it measures and measuring are different things — and I've been a better reviewer for it ever since. Dodged that bullet, honestly.
So here's my parting challenge: next time, don't open with “which Fluke multimeter should I buy.” Open with the signal you're chasing, the smallest change you genuinely need to see, and the environment you'll be measuring in.
“Work backward from the measurement, not outward from a brand.”
I keep that line pinned to the top of our internal purchasing guide. The spec sheet — not the logo — is the contract you want to sign.
That's the quality bar I apply to every unit that crosses my bench, whether it's a $200 used meter or a $20,000 order. It'll work for yours too.
