The best heater for a 20-gallon tank

The sum, for this exact tank

20 US gal rated, less 15% for substrate and fill line, is 17 gal of water. A 78F target minus a coldest-room 68F is a 10F lift. 17 x 0.35 x 10 = 59.5W required. Nothing is sold at 59.5W, so the answer is the 75W.

A rated 20-gallon holds about 17 gallons of water once there is substrate in it and you have stopped filling below the rim. Against a typical American room - Aquarium Co-Op publishes 67 to 69°F as common - and a 78°F target, that is a 10-degree lift and 59.5 watts of heat.

Seventy-five watts is the next size sold. It is also, unusually, inside the manufacturer’s own recommendation rather than at the edge of it: EHEIM publishes its 75W as suiting aquariums of about 60 L with a 100 L ceiling, and a US 20-gallon is 75.7 L. That is the only product on this site where the maker’s recommended volume and our independent sum agree on the same tank.

How it changes with your room

Read down the last column. At 70°F and above the sum is satisfied by a 50W; from 68 down to 66 it wants 75; at 64 it crosses into 100. The 68°F row is the one most American homes live on, which is why 75 watts is the headline answer.

At the 64°F row you are asking for a 14-degree lift, past the 12°F Fluval publishes as the ceiling for a single heater. The arithmetic still resolves - 83.3W required against 100 available - but at that point we would rather you warmed the room, fitted a lid, or split the load than put 100 watts behind one thermostat.

What we ranked these on

We have not run these heaters. No test bench, no drawer of samples, so every criterion below can be checked from a published document rather than taken on trust.

Published spec depth. Wattage is on every box. Rated or recommended volume is on most. The third figure - a maximum lift over ambient, or a stated control accuracy - is what tells you whether the heater works in your room, and almost nobody publishes it.

Sizing honesty. Whether a buyer can compute their own answer from what the maker publishes, or is asked to trust a volume range with no conditions attached.

Cold-room headroom. Behavior at a 64°F room rather than a comfortable one. At twenty gallons this is the criterion that separates the 75 from the 100.

Concealability. A 20-gallon is roughly 24 inches long and 12 to 16 tall. That is more back glass to hide something along than a nano gives you - but the EHEIM 75 is 270 mm and the 100 is 319 mm, and those are real objects in a display.

Control. Preset versus a dial. At twenty gallons you are more likely to be running a community than a single betta, and a community is the case where being able to move the set point actually matters.

Why the EHEIM and not the Fluval P range

Because the P range stops. Fluval publishes the P50 at 50W rated to 15 gallons, and a 20-gallon is past that. The honest options at this size are a larger adjustable heater or two P50s, and the first is simpler.

EHEIM is also the only manufacturer in this class that publishes a control accuracy - plus or minus half a degree Celsius - rather than only a volume range. That is a different and more useful disclosure than Fluval’s maximum-lift note, and at a tank size where you may be holding a specific temperature for a specific species it is the figure worth having.

What EHEIM does not publish is a maximum lift, so its volume range cannot be read as a cold-room guarantee. Our sum is what fills that gap, which is exactly why we publish it.

One 75W, or two 50s?

Twenty gallons is where splitting the load stops being a thought experiment. Two 50W heaters at opposite ends of the tank give you 100 watts of capacity with two independent thermostats, and that changes both failure modes for the better:

• A thermostat that sticks on has half the power to work with, and 17 gallons of water to dump it into. That is a survivable overnight rather than a fatal one.
• A thermostat that dies leaves you at half heat rather than no heat, which in a January bedroom is the difference between a cool tank and a cold one.
• Heat distributes better across 24 inches from two points than from one.

What it costs is two cords, two heater bodies in the display, and roughly double the money. We would take the single 75 in a normally heated house and split in a room that genuinely runs cold - the same conclusion the safety page reaches from the failure-mode side.

Placement in a 24-inch tank

Horizontally, low, along the back glass, with the filter return moving water past it. Heat rises, so a heater lying near the substrate puts its output into the coldest water and lets the flow distribute it; one standing vertically in a corner warms the top third and leaves a cold layer where the fish sleeps.

Then put the thermometer at the opposite end of the tank. Not beside the heater, where it reads the heater rather than the tank. In a 24-inch tank that distance is long enough that the two readings genuinely differ, which is itself useful information about whether your flow is distributing heat.

What twenty gallons changes about the whole tank

Thermal mass is the quiet advantage nobody puts on a box. 17 gallons of water takes hours to move a degree when the power fails, where a nano moves in minutes. It is the same reason the filter decision changes at this size - more volume means turnover stops being the binding constraint - and it is the strongest argument for a bigger tank being easier, not harder, than a small one.

What we cannot tell you

Which of these holds its calibration after a year, which brand sticks on most often, or whether EHEIM’s published half-degree accuracy survives two winters in a real room. Those are the questions that decide whether a heater is any good, they need a lot of heaters and a lot of time, and we would rather say so than dress a specification sheet up as a test report.