How to Read Multimeter CAT Ratings and Fuse Ratings

If you take one thing from this guide: the CAT IV 600V rating printed on a multimeter's faceplate is decided mostly by the voltage rating of the few fuses inside the case, not by its accuracy. I use four currently sold models to show this: the Fluke 87V, Klein Tools MM720, Klein Tools MM420, and UNI-T UT61E. Every ASIN was confirmed by actually loading the Amazon product page before I wrote this, not guessed from a model number. Prices and stock status are as of October 12, 2026, and where sources disagree I show the disagreement instead of picking one number.

This post contains affiliate links. If you buy through them I may earn a commission, and it does not change your price.

TL;DR: fuses first, CAT rating second, counts last

• A CAT rating is a withstand rating, not an accuracy rating. CAT III covers upstream of fixed installation, CAT IV covers the service entrance and meter base side, where transient energy is higher. CAT IV 600V is genuinely stricter than CAT III 600V, not just a bigger number.
• The CAT rating is ultimately cashed out by the fuses' voltage rating and breaking capacity. An independent teardown found the UNI-T UT61E+ carries a 1000V-class CAT marking on the front while its 1A and 10A fuses are rated 240V, and the PCB silkscreen says 240V too.
• Counts only decide how small a change you can see, not whether the reading is accurate. The UT61E's 22,000 counts can display 11.983V. Its basic accuracy on that range is still ±(0.05%+5). Resolution and accuracy are separate things.
• The four routes — Fluke 87V (industrial and motor-drive work, low-pass filter, roughly $446–485), Klein MM720 (LoZ mode, the ghost-voltage cure, roughly $95–100), Klein MM420 (around $60, best seller of the group but CAT III 600V only), UNI-T UT61E (22,000 counts plus PC data logging, out of stock and with a disputed CAT claim).
• The Amazon Basics multimeter (DT-9561, ASIN B083V7N12H) returns 404 on amazon.com. I do not link it and do not count it in the four. It is the one product I wanted to recommend and cannot.

Why the fuse matters more than the accuracy spec

When a multimeter is measuring voltage and the circuit is miswired or hit by a transient, current has to be interrupted by a fuse. The fuse's voltage rating determines how much voltage it survives without failing; its breaking capacity determines how large a fault current it can safely interrupt. The CAT rating on the faceplate is a promise about the whole assembly protecting the operator in the stated environment, and the weakest link in that assembly is often a fuse that costs a few cents.

There is a textbook counterexample. Independent teardowner Kerry Wong documented the UNI-T UT61E+: both the 1A and 10A fuses are rated 240V at 6mm × 25mm. By contrast UNI-T's actually certified UT161E uses a 600mA 1000V fuse on the low-current range (6mm × 32mm) and an 11A 1000V fuse on the 10A range (10mm × 38mm). The same write-up notes the UT61E+ carries no UL or CE marks on the meter itself.

Two independent conclusions on the EEVblog forum are worth quoting. One is that the fuses are only 250V and lack the breaking capacity a real CAT IV 600V meter needs. The other questions the certification step directly: Intertek would not grant approvals beyond 61010 in that situation.

Put those together and the conclusion does not need hedging: when a meter's CAT marking disagrees with the voltage rating of its own fuses, believe the fuses, and treat the CAT rating as an unverified claim rather than an established fact. That is not timidity. The faceplate is marketing copy; the fuse silkscreen is a component spec.

It is worth being precise about where CAT boundaries sit. The Fluke 87V product page states that all inputs are updated to the third edition of EN 61010-1 to CAT III 1000V and CAT IV 600V, and that the meter is designed to withstand spikes above 8000V. CAT III 1000V and CAT IV 600V appearing together is not redundant labeling: the first corresponds to fixed installation at the panel, the second to the line side closer to the service entrance, and the two positions see different transient energy.

The four routes: safety class, resolution, field readability

Fluke 87V — the industrial and motor-drive route

The Fluke 87V Industrial Digital Multimeter, ASIN B0002YFD1K, was confirmed to render a real product page before publication, rated ★4.8 from 862 ratings, ranked #41,886 in Tools & Home Improvement.

Its core capability is a trio: a Fluke-designed low-pass filter for frequency and voltage measurement on adjustable speed drives, Peak Capture that grabs transient spikes as fast as 250 microseconds, and switchable display resolution between 6,000 and 20,000 counts. It ships with TL75 test leads, AC175 alligator clips, and an 80BK temperature probe, runs on four AA cells, weighs 1.3 pounds, and carries a limited lifetime warranty.

Real defects. It is heavy. At 1.3 pounds in a 9.49 × 6.5 × 3.5 inch body, it is a bulk item in a tool bag, not something you pocket. Reviewers consistently single out the included TL75 leads as the weakest part of an otherwise serious tool: the insulation stiffens in cold weather and the probe tips are not as sharp as aftermarket replacements, and most owners swap to silicone leads within the first month. The battery door, redesigned only in 2024, still uses an older wire-lead style that feels dated next to current snap-fit designs. If you never touch motor drives or three-phase power, a large share of this meter's capability will sit idle.

Price sources conflict, shown as-is. camelcamelcamel records $446.15 sold by Amazon as of 2026-10-09 and $444.75 for third-party new as of 2026-10-08; the Amazon product page the same day showed $484.53 against a $521.99 list price; the price-comparison site budgetbuzz recorded $468.82 on 2026-09-28. The spread comes from timing and channel, not from one of these being transcribed wrong. Check the page yourself before buying.

Klein MM720 — the low-impedance route built for ghost voltage

The Klein Tools MM720 Digital Multimeter, ASIN B0B57PDQJC, rendered a real product page, ★4.7 from 403 ratings, ranked #27,506 in Tools & Home Improvement. Specs are 1000V AC/DC, 10A AC/DC, 60MΩ resistance, dual CAT III 1000V and CAT IV 600V marking, IP42, 2-meter drop rating, 345 grams, two AAA cells.

It earns its own slot because of LoZ (Low Impedance) mode. When a dead wire runs parallel to a live one, a meter's own high input impedance will faithfully display the few volts picked up by capacitive coupling. The wire is open, yet the meter insists on 60 to 100V. This ghost voltage sends people hunting for a short that does not exist. LoZ mode applies a deliberate controlled load so the phantom reading collapses; if the number holds steady, the voltage is real.

Real defects. This is the product whose stock status most needs disclosure. On the day I fetched it, the Amazon page showed no featured offer available with other sellers only, meaning no Amazon-direct offer and third-party fulfillment only. camelcamelcamel records $99.97 Amazon-direct as of 2026-08-27 and $95.00 for third-party new as of 2026-09-18. Confirm the seller and shipper yourself before buying; third-party stockouts or cross-border shipping change the delivery timeline substantially. Two ergonomics costs as well: at 345 grams and over 10 inches long it is the bulkiest meter in this group, and two AAA cells do not last a heavy field week. The magnetic hanger (Cat. Nos. 69445 or 69417) is a separate purchase. It also has no µA or mA ranges and no PC or Bluetooth logging whatsoever.

Klein MM420 — the $60 tier, best seller but a safety class lower

The Klein Tools MM420 Digital Multimeter, ASIN B0B57PFFYX, rendered a real product page, ★4.5 from 756 ratings, ranked #5,621 in Tools & Home Improvement — the best seller of the four. camelcamelcamel records $64.69 Amazon-direct as of 2026-10-11, $59.99 third-party new as of 2026-10-10, an all-time low of $45.68 on 2024-06-23, and a $64.98 list price. Specs are 600V AC/DC, 10A, 50MΩ, TRMS auto-ranging, CAT III 600V.

Real defects, and they sit on the safety boundary. It is CAT III 600V with no CAT IV. By the definition of the categories, that rules it out for work at the service entrance side. It also lacks LoZ, so the ghost-voltage problem the MM720 solves is present here unchanged: if phantom readings show up while tracing old wiring in a renovation, this meter cannot help. What you get in exchange is LED lead-alert indication when probes land in the wrong jack, and a body more compact than the MM720. The trade is explicit: for 220V household circuits and board-level repair this is enough and it is cheap; the moment you touch the line side, saving those $40 is the wrong call.

UNI-T UT61E — the 22,000-count route, with asterisks

The UNI-T UT61E Digital Multimeter, ASIN B06XSYZZBV, rendered a real product page, ★4.6 from 90 ratings. The page's spec table lists 1000V DC accuracy from ±(0.1%+2) on the UT61A/B/C/D up to ±(0.05%+5) on the UT61E, a 20A current range, 40MΩ to 220MΩ resistance, a 22,000-count display, True RMS, and PC connectivity.

A model-name trap worth stating plainly. The ASIN I verified is the UT61E, while the teardown and forum discussion cited in section one concern the UT61E+, the 22,000-count unit. The names look nearly identical but they are different meters, and the fuse conclusions from the forum apply directly only to the UT61E+. The same product page's comparison table lists UT61E and UT61E+ side by side, so if you are carrying a conclusion over, match the model on your own unit first.

Real defects. First, stock and price are both unreliable: camelcamelcamel shows all three price types as out of stock for this ASIN, last tracked 2023-08-03, with no Amazon-direct price ever recorded and only a third-party new range of $49.98 to $127.99 against a $63.99 list price. The #2,026,789 category rank confirms it barely moves on the US store. Second, the safety claim is disputed, and the third-party sources do not even agree with each other: hobbyistscope records CAT II 1000V and cites teardown evidence that the internal fuses do not meet the claimed class, while another review states CAT III 1000V / CAT IV 600V. Whichever you accept, you need to check the body label and the fuses yourself. Third, input impedance dropped across the generation change: Kerry Wong records the UT61E+'s DC mV range falling from the UT61E's 3GΩ to 1GΩ, with the AC mV range at only 10MΩ against the original's 3GΩ. Fourth, the range selector was found to stick between two positions during testing, and the PC software would not connect on Linux, with the literal error text "Cannot enable. Maybe the USB cable is bad?"

Amazon Basics DT-9561 — the one I wanted to recommend and cannot

This section exists because it defines the credibility boundary of the whole piece. The Amazon Basics Digital Multimeter (model DT-9561, 6000 counts, True RMS, CAT IV 600V, IP67, type-K thermocouple, 9V battery, 255g, ASIN B083V7N12H) returned Page Not Found on amazon.com across repeated fetches before publication, with a 2296-byte response body. It is delisted on the US store. Pages still exist on amazon.de and amazon.co.uk, so this is not a product that never existed; it is a product you cannot buy on amazon.com.

So it gets no link here and does not count as a fifth route. If you can buy it in your market, the spec combination is still reasonable for the price: 6000 counts plus True RMS plus CAT IV 600V plus IP67 is unusual under $100, and the recurring third-party complaints are a quiet continuity buzzer and auto-ranging that pauses before settling on a reading. Silently dropping a same-tier product out of a four-way comparison would be worse than stating plainly that it is gone.

Putting accuracy back where it belongs

With safety class settled, three numbers still need to be kept apart.

Counts determine resolution. A 22,000-count meter on the 2.2000V range resolves to 0.1mV, so you can see whether a 3.3V rail reads 3.298V or 3.301V. A 4,000-count meter shows both as 3.30V. The UT61E product page's spec table spells the progression out: 4,000 counts on the UT61A/B/C, 6,000 on the UT61D, 22,000 on the UT61E and UT61E+.

Basic accuracy determines whether the reading can be trusted. It is written as ±(0.05%+5): the percentage term scales with range, and the 5 is a fixed count added at the least significant digit. On low ranges that fixed term can outweigh the flattering percentage. This is why "22,000 counts" and "high accuracy" are not synonyms.

Input impedance determines how much the meter distorts what it measures. The same teardown recorded the UT61E+ DC mV range falling from 3GΩ to 1GΩ. That is invisible on most bench work, but on sensor signals or high-impedance nodes it lands in your error budget.

Laid out side by side: 22,000 counts make a reading look fine-grained, ±(0.05%+5) determines whether you can believe it, and the fuse voltage rating determines whether it survives the mistake. None substitutes for another.

A decision tree with three branches

Branch one: you need the line side, mains trunk, or a distribution board. Anything at CAT III 600V or below is out, which leaves the 87V and the MM720. On a budget take the MM720 (around $95–100) and confirm both the CAT IV 600V marking and that Amazon itself is the seller. If you need adjustable-speed-drive frequency measurement, Peak Capture, or the lifetime warranty, take the 87V ($446–485) and accept the 1.3-pound weight and the leads you will replace.

Branch two: you are at a bench debugging boards and need to see small voltage changes. Prioritize counts and true RMS. The UT61E's 22,000 counts genuinely lead here, but you accept three things: no Amazon-direct stock, a CAT claim whose input protection you must verify yourself, and PC logging officially supported only on Windows. If you would rather not carry those, the MM720's 1000V range plus True RMS plus CAT IV is the steadier compromise.

Branch three: you just want to know whether a wire is continuous or an outlet is live. The MM420 ($64.69 Amazon-direct, $59.99 third-party) is enough, and it is the best seller of the four. Giving up CAT IV and LoZ is reasonable when you are not doing service-entrance work. But keep ghost voltage in mind: when a dead circuit in an old building reads tens of volts, that is far more likely a phantom reading than real power.

Troubleshooting: four failure modes I would actually hit

Mistake one: treating three consistent readings as accuracy. Reading the same meter three times in the same spot will agree, which proves the meter works, not that it is right. Verifying accuracy requires a known source, such as a reference voltage source of known precision or an already-calibrated meter to compare against.

Mistake two: plugging leads in before selecting the range. Seating probes in the wrong jack while the circuit is live is the most common and most avoidable way to destroy a multimeter. Both the 87V and the MM420 alert you with an LED or buzzer on misplacement (Lead Alert / Input Alert), and the MM720 carries the equivalent lead-alert design. Of everything in these four meters, that is the feature I would rank highest — worth more than an extra range.

Mistake three: letting the CAT marking stand in for a safety judgment. Back to section one: the CAT label is a promise, the fuse silkscreen is a fact. If you will be doing high-voltage work where nobody is checking over your shoulder, the only responsible approach is to select on the most conservative rating available and to open and inspect any meter of uncertain provenance.

Mistake four: selling a cheap meter's high count as precision. All four mistakes orbit one root cause: resolution, accuracy, and safety are three orthogonal dimensions, and treating one label as the whole conclusion is the most expensive mistake in multimeter shopping.

FAQ

Q1: What is the actual difference between CAT II and CAT III?

CAT III applies to fixed installation, such as a fixed branch circuit inside a panel. CAT IV applies at the service entrance, on the line side, and immediately around the meter base. The latter sits closer to the supply and sees higher transient energy, which is why the requirements are stricter at the same voltage. The short version: a CAT IV meter can do CAT III work, not the other way around.

Q2: Does a higher count always mean a better meter?

No. Counts only raise resolution. What determines trustworthiness is the percentage and fixed-count terms in the basic accuracy spec, plus whether the input impedance suits the circuit. A 40,000-count meter rated ±(1.0%+10) can be worse in practice than a 4,000-count meter rated ±(0.05%+2).

Q3: Can these meters measure temperature?

Only the Fluke 87V bundles an 80BK probe covering -200°C to 1090°C out of the box. The MM720 and MM420 both list temperature measurement and ship a thermocouple, while the UNI-T UT61E lacks it, which multiple third-party reviews list as a genuine shortcoming. If you routinely check heatsinks, motor bearings, or case intake temperature, confirm the model has a temperature channel before buying.

Q4: Is True RMS actually necessary?

Only when the signal is not a pure sine wave. Switch-mode power supplies, variable-frequency motor drives, LED dimmers, and PWM audio all produce distorted waveforms. An average-responding meter reads these low and sends your entire diagnosis in the wrong direction.

Q5: How do I confirm a ghost voltage?

Use a meter with LoZ mode such as the MM720: switch to LoZ and if the reading collapses toward zero it was capacitive coupling; if it holds steady, the voltage is real. Without LoZ you can put a resistor of known value in parallel across the point under test and watch whether the reading is pulled down. It is a crude method, but the conclusion holds.

Sources and update cadence

Every ASIN was confirmed by loading the actual Amazon product page before publication rather than guessed from a model number. Ratings, review counts, and category ranks come from the page as fetched; prices come from camelcamelcamel and the product pages, with the Fluke 87V source conflict shown in the body. CAT classes, fuse voltage ratings, input impedance, and true-RMS behavior were cross-checked across manufacturer spec pages, Amazon product page descriptions, an independent teardown, and the EEVblog forum.

Multimeter prices and stock move far more than keyboard and mouse prices. Treat these four tiers as needing a recheck within six months.

👉 Join Xiaomi MiMo Platform: Leading AI model platform with cost-effective inference

👉 Join Aliyun AI: Top AI products with exclusive coupons for business innovation

📌 This article was AI-assisted generated and human-reviewed | TechPassive — An AI-driven content testing site focused on real tool reviews