Skip to content
Beam & Bezel

EDC Flashlights

What to look for when buying a flashlight

Six specifications decide whether a flashlight is any good. Peak lumens, the one on the front of the box, is fifth.

By Scooter M. · Published · Updated

What to look for when buying a flashlight — ranked picks, with live prices
#ProductBest forOutputPrice
1
No
image
Streamlight MacroStream USBTop pick500 lumens, 2,000 candela, 90 m and two hours on high — every figure published per mode by the maker. That is the reason it leads.
The best-documented small EDC light here500 lm
2
No
image
Wurkkos FC11C1,200 lumens at a stated CRI of 90 for a budget price. Color rendering is the spec that changes how a light feels, and this is the cheapest way to get it.
High color accuracy for the least money1,200 lm
3
No
image
Fenix PD36R Pro2,800 lumens from a 21700 cell holding 18 watt-hours, at 169 grams. The biggest energy store in the tactical set.
The most output and energy in a tactical body2,800 lm

We earn a commission if you buy through the links on this page, at no extra cost to you. It does not change what we recommend or the order they appear in. Full disclosure. We show a live price when we have one and the words “Check price” when we do not. We never print a number we cannot currently verify.

A flashlight and small tools laid out on a dark surface
Licensed stock photograph, for illustration. It is not a photograph of any product on this page and we have not handled any of them.

There is a reliable way to tell a good flashlight listing from a bad one before reading a single review: count how many specifications it publishes. A maker who has measured their light tells you six numbers. A maker who has not tells you one, and it is always the biggest one.

1. The cell, and what it holds

This is the specification that decides everything downstream. Convert whatever the maker publishes into watt-hours: milliamp-hours divided by 1,000, times the nominal voltage.

What each format holds
CellWatt-hoursVerdict
1 x AAA alkaline1.3Keychain lights only
1 x AA alkaline1.9Long low-mode runs; poor at high output
163402.9High output in a very short body; short runtime
1865012.2The default for a reason
2170018.0Best runtime, fattest tube

2. Sustained output, not peak

The peak figure lasts until the driver steps down to protect the emitter, which in a pocket light is usually between 30 and 120 seconds. What matters is the level below that, which the light will hold for hours.

A good listing publishes a per-mode table with output and runtime for each level. If it publishes only a turbo figure, it is telling you which number is the flattering one. Flashlight runtime explains how the standard defines the measurement.

3. Candela, or a beam distance you can convert

Lumens tell you how much light there is. Candela tells you where it goes. Two lights of identical output can differ by a factor of ten in reach depending on the reflector.

4. The interface

The best specification sheet in the world does not help if you cannot get to the mode you want. Things worth checking: whether the light starts in low or in turbo, whether there is a lockout, whether the mode order is memorable, and whether the switch can be found by feel in the dark.

A light that starts in turbo will blind you every time you switch it on indoors. A light with mode memory returns to where you left it, which is almost always what you want.

5. Peak lumens

Fifth, because in a pocket-sized light it is the specification most limited by physics and least connected to use. Somewhere between 500 and 1,000 peak is plenty for anything a handheld light does at close range.

Beyond that you are buying a burst, and the burst is limited by how much heat the body can shed, not by the emitter. This is why a metal light and a plastic light of the same nominal output behave differently after two minutes.

6. Ingress rating and impact resistance

IPX4 is splash, IPX7 is immersion to a meter, IPX8 is continuous immersion, and IP68 adds dust tight. Impact resistance is quoted in meters and tested by the ANSI FL 1 protocol. Both are meaningful specifications that many makers simply omit.

The listing red flags

  • Five-figure or six-figure lumen claims. A genuine 100,000-lumen light draws kilowatts. The figure is not a measurement of anything.
  • A single output number and nothing else. No candela, no runtime, no weight, no rating.
  • A runtime with no mode attached. 'Up to 100 hours' on a light rated 2,000 lumens is a low-mode figure and it is being quoted next to the wrong number on purpose.
  • 'Military grade' or 'tactical' with no specification. Neither term has a definition.
  • Zoom on a cheap light. The mechanism loses output, produces a square-edged beam and is the first thing to fail.

The picks, in full

No
image

Top pick · The best-documented small EDC light here

Streamlight MacroStream USB

500 lumens, 2,000 candela, 90 m and two hours on high — every figure published per mode by the maker. That is the reason it leads.

2,000 candela and 90 m are consistent with each other, which is a better sign than it sounds. The FL1 beam distance is the square root of candela divided by 0.25 lux: the square root of 2,000 over 0.25 is 89.4 m. Streamlight's published figures check out to within a meter.

Two hours on high for 500 lumens implies roughly 2.5 watt-hours of usable energy, which is about right for a light this size. Nothing here is being oversold.

62 grams is pocketable in a way that a 21700 light is not, and that is the trade you are making against a bigger cell.

What works

  • Every headline figure published per mode, and internally consistent
  • 62 g with a real clip
  • The arithmetic between its candela and its distance actually reconciles

What does not

  • IPX4, so splash-resistant and no more
  • Built-in cell rather than a replaceable 18650
  • 500 lumens is modest against the 1,600-lumen headlines nearby

Skip this ifyou want reach. 2,000 candela is a close-range beam and no setting will change that.

No
image

Pick 2 · High color accuracy for the least money

Wurkkos FC11C

1,200 lumens at a stated CRI of 90 for a budget price. Color rendering is the spec that changes how a light feels, and this is the cheapest way to get it.

CRI 90 versus the CRI 70 typical of a cheap cool-white emitter is the single most noticeable upgrade in a flashlight, and it is nothing to do with brightness. Reds stop looking brown. Wire insulation colors become distinguishable.

The cost of high CRI is output: the same emitter family loses roughly 10 to 20 percent of its lumens for the better phosphor. A 1,200-lumen high-CRI light is doing well.

18650 is the right cell format here — replaceable, standard, and enough energy to hold a useful level.

What works

  • CRI 90 at a price where CRI is usually not mentioned at all
  • Standard 18650, replaceable anywhere
  • USB-C on the light itself

What does not

  • No sourced runtime, candela, beam distance or IP rating
  • High CRI costs output; treat the 1,200 figure as generous
  • A budget brand with a thin published spec table

Skip this ifyou need maximum reach per dollar. High-CRI emitters throw less than the cool-white ones they replace.

Published specification

Peak intensity
not published
Beam distance
not published
Runtime
not published
Weight
not published
Ingress
not published
No
image

Pick 3 · The most output and energy in a tactical body

Fenix PD36R Pro

2,800 lumens from a 21700 cell holding 18 watt-hours, at 169 grams. The biggest energy store in the tactical set.

A 5,000 mAh 21700 at 3.6 V is 18.0 watt-hours — nearly three times what the ProTac HL-X carries. That is the case for the 21700 format in one number.

380 m of reach on 2,800 lumens works back to about 36,100 candela, almost identical to the ProTac's 35,000 from 1,300 lumens. The Fenix is spending its extra output on spill, not on distance.

169 g including the cell is only 7 g more than a ProTac on CR123A, for nearly triple the energy. That is what four years of cell development buys.

What works

  • 18 Wh of stored energy, the most in the tactical set
  • 2,800 lumens with 380 m of published reach
  • Standard 21700 cell, USB-C on the light

What does not

  • No published IP rating on the page we read
  • No published candela, so the throw figure cannot be cross-checked
  • 2,800 lumens will step down quickly in a body this size

Skip this ifyou need a shelf-stable light. A rechargeable cell self-discharges; for a drawer light, CR123A primaries are the right answer.

Published specification

Peak intensity
not published
Runtime
42 h maximumLowest mode; Fenix publishes the full per-mode tableFenix PD36R Pro — manufacturer spec page (retrieved September 1, 2026)
Ingress
not published(Not stated on the page we read)

Questions people actually ask

What is the most important flashlight specification?

The cell, converted to watt-hours: milliamp-hours divided by 1,000, times nominal voltage. It decides the runtime, and the runtime decides whether the light does the job. Peak lumens is fifth on the list.

How do I know if a flashlight's lumen claim is real?

Check whether the maker also publishes candela and beam distance, and whether the two reconcile: beam distance is the square root of candela divided by 0.25. When those agree, the figures are almost certainly measured. When only a lumen number exists, you have nothing to check.

How many lumens do I actually need?

Indoors, 50 to 200. Outdoors at close range, 200 to 500. For genuinely searching at distance, the number that matters is candela, not lumens — a 600-lumen thrower reaches further than a 2,000-lumen flood.

What does 'tactical' mean on a flashlight?

Nothing with a definition. In practice it usually means a tail switch, a strobe mode, a crenellated bezel and an aluminum body. Judge the light by its published specifications rather than by the word.

Is a metal flashlight better than a plastic one?

For sustained high output, yes — aluminum sheds heat, so the light steps down later and less. For impact resistance, polymer often wins: the Streamlight PolyTac's published 3 m impact rating is the highest in our set.

What is thermal step-down?

The driver reducing output to keep the emitter and the body within safe temperatures. Every high-output light in a small body does it, usually within 30 to 120 seconds. It is why sustained output matters more than peak.

Sources

Every figure on this page came from one of the documents below, read on the date shown. We do not cite other review sites, forum posts or video reviews as sources of fact.

A night sky over a dark landscape

Lumens vs candela

The distinction that decides reach, with the conversion formula and a table of real published figures.

Found a number that is wrong? Tell us and we will fix it and say so. Our methodology shows every formula on this page.