TL;DR
If your battery dies in 2–3 days of parking, either you have:

1. A weak/aged battery.
2. Some charging problem.
3. Excessive key-off (parasitic) draw.

Be sure to wait for your vehicle to get fully into “sleep” mode before making any judgment about draw–this can take anywhere from 10 to 45 minutes, and occasionally longer.
Only after that can you measure current draw properly (don’t crank the engine with the meter on, or you’ll fry it) and then isolate the fault, either by pulling fuses or by the cleaner “voltage drop across the fuses” method.

Often the source of excessive key-off draw are glovebox lights, lights in the trunk or by the battery, aftermarket dash cam/amps/alarms, stuck rad fan relays, infotainment modules not properly going to sleep, and alternator diodes leaking.
After you fix the problem, reread the key-off draw and keep your battery topped off; deep-discharging can permanently weaken it.

How a battery dies in 2–3 days (and why jumping it isn’t a fix)

A modern vehicle is never really “off”. Clocks, security systems, computers, and telematics all pull some amount of current 24/7. That’s normally fine, but when something stays awake (or a light/relay stays on) it can drop the battery below starting voltage in a few days. It’s also common for a marginal battery to appear “fine” after a charge, then fail quickly once parked, particularly in cold weather or with more than one deep discharge. This is why the best method is to: confirm battery/charging basics, then isolate excessive key-off draw.

Before you chase milliamps: 10 min Triage

• Battery age: If it’s ~3-5+ years old (lead-acid actual life varies considerably), keep “weak battery” on the suspect list even if parasitic draw is found.
• Connections: Make sure terminals are clean, tight, and grounds secure. A bad connection can mimic “battery drain” symptoms.
• Charging: If the vehicle always needs a jump even after running, take the alternator/charging system in for testing (most parts stores and repair shops can do this quickly).
• Behavior clue: Dies 2-3 days parked frequently implies moderate draw (hundreds of mA) or weak battery plus small draw. Dies overnight could easily indicate larger draw (amps), and/or stuck light.

If your battery has been fully discharged multiple times, assume it may be damaged even after you fix the draw. Plan to recharge it fully and then load-test it.

What counts as “normal” parasitic draw?

There isn’t one number to shoot for because features and module strategies vary by vehicle. However, many sources regard roughly 50 mA as typical goal after the car is all the way asleep, and >100 mA as “suspicious” amount in many models. Some newer vehicles will legitimately have a higher number (eg ranges of ~50-85mA are often said). The only correct spec is found in your vehicle’s service information.

Tools you’ll want (and why they matter)

• Digital multimeter with a fused 10A input (and mA range): You’ll want to start on the high range to protect the meter, and switch once you know you’re in the right ballpark.
• Optional: DC clamp meter with milliamps resolution: Good for intermittent draws, avoids removing battery. Watch for cheap clamps that doesn’t have enough resolution on small draws.
• Fuse puller and flashlight / headlamp: You’ll be in fuse panels for a while.
• Painter’s tape / latch tools: To tape door/hatch/hood switches “closed” while panels are open. How to take control of Mr. Key-off Draw step 8: Keep reading for the whole process!
• Wiring diagrams/service info: Important once you find which fuse is draining, as one fuse can power several modules.

Out of Order: measure key-off draw (the right way without waking the car)

1. Charge the battery, if necessary. Testing for parasitic draw with a low battery sometimes leads to confusing results (though not always—some modules will “wake up” when they see a low voltage).
2. Prepare the car to “go to sleep”: Put the key in the off position, turn the lights off (interior and exterior), and remove anything plugged into 12V sockets and USB ports, and make sure the key fob is not too close to the vehicle (it might wake up if the fob is near).
3. Simulate closed doors, trunk, hood, etc., and your job will be easier. Tape over the pinswitch at each door or latch it closed on top of the striker (be careful not to let it be accidentally full-slamming into the latched striker).
4. Wait for the car to go to sleep, which can take 10–45 minutes, and sometimes even 30-120 minutes, depending on the modules. You can’t rush this (if you try and measure too early, you’ll notice the greatly, greatly elevated draw, but it’s not the problem). Wait for sleep mode, and then let’s move on.
5. Do not go directly to step 6! Start on the highest current range (probably 10A) and put the leads in the correct jacks.
6. Connect the meter in series with the battery negative cable: battery post to meter to the removed negative cable. Make sure that connection is solid (if it opens, modules can wake and you’ve got to start waiting for the car again).
7. See what the draw is with the car asleep: it’s normal to see a higher draw at first, then a downstep, then record the steady value at ‘fully asleep’.
8. If the ‘asleep’ value looks excessive, go to isolation. Normal? Could be battery health problem or an intermittent draw that isn’t there right now.

Isolation method 1 (classic): pop fuses and watch it drop

This is the old school approach: meter still in series, car still fully asleep, pop one fuse at a time and look for a drop. When it drops, the last fuse you pulled is (probably) feeding that circuit.

1. With the car still fully asleep and current steady, start with interior fuse panel fuses (usually body functions) and then under hood.
2. Pull ONE fuse, wait a few seconds, see if the current changes. Be sure to reinstall it before pulling the next, unless you have a pretty decent hunch that it’s the bad guy and you’re ready to go in.
3. If the car wakes up (current spikes higher and stays high), stop, let it go back to sleep before proceeding.
4. Once you find a fuse that ‘drops’ a significant amount, write it down and look up everything that fuse powers using wiring diagrams/service info. Then isolate within that circuit: unplug modules, disconnect relays or aftermarket devices on that circuit one at a time until the draw drops to normal.

Isolation method 2 (and many times better): voltage drop across fuses (no fuse pulling)

Another popular modern technique is to measure the millivolts (mV) drop across each fuse. Any fuse that is carrying some current will have a small voltage drop across it. Fuses that read 0.0 mV means there is essentially no current flowing; this speeds up the process as you won’t have to pull a fuse to determine if it is “active,” which might awaken many modules in the process. Here’s how to do it:

1. As with the above methodologies, allow the vehicle to “sleep” first (rules as above).
2. Set your multimeter to millivolts (mV DC), and touch the meter probes to the two exposed test points on top of each blade fuse. Most modern blade fuses have these.
3. Log down any fuses that show a measurable mV drop; any that read 0.0 mV are not carrying any current at that moment.
4. Convert that mV drop with a fuse voltage-drop chart (these vary by fuse type and fuse rating). Service info providers will show these and methodologies.
5. Once learning that a fuse/circuit has an unusually high current draw on it, switch to component level isolation (doing your unplugging of relays/module/accessories on that circuit).

Don’t forget the “no-fuse” culprits (these can bypass your fuse panels)

Some of the most common drains aren’t in the obvious fuse-by-fuse path, or they’re just easier to check upfront before you sit for an hour working the fuse box. Two big categories: alternator diode leaks and age old aftermarket add-ons wired straight to the battery.

Quick alternator check (key-off drain test)

1. With the parasitic draw clearly excessive and the vehicle asleep, note the current reading.
2. Disconnect the alternator from its electrical connections, and do so in a prudent fashion so you don’t accidentally short the B+ terminal to anything else with your wrench.
3. Recheck key-off draw. If it drops substantially, you have a likely alternator diode leak; verify this with proper charging and/or diode tests or refer for professional help.

Aftermarket accessories (the repeat offenders)

• Dash cams in “parking mode” but no low-voltage cutoff
• Amplifiers and subwoofer “remotes” wired to constant power, without a good strategy on remote turn-on
• Remote start and other aftermarket alarm systems
• OBD-II dongles plugged in and connected 24/7
• Trackers/fleet devices added by the last owner or lender

Plug at that fuse: how to pin down exactly what you’re dealing with

Finding the fuse is only the first “yes/no” determination; the real work is in isolating just what it is that is either still powered, or at least waking the car back up on that circuit. Use this workflow to avoid random parts swapping.

1. List everything on the circuit. Wiring diagrams/service info should identify everything on the fuse load side (modules, relays, lamps, sockets, etc.)
2. If all else fails, divide and conquer. If the fuse feeds a relay, pull the relay first. If the drop drops, you just narrowed it to the relay contacts or the load side of that relay.
3. Unplug loads one at a time. Modules usually have connectors you can unplug; lamps have sockets; aftermarket items have inline fuses. Each time you disconnect something, watch for a current drop (or reduced mV drop at that fuse).
4. Check ‘simple’ switches first: glovebox, trunk, and underhood lights are famous because the bulb might be hidden plus the switch might fail or be misadjusted.
5. Inspect for heat/noise. A stuck relay, or motor circuit might have that relay warm, or you might hear a faint hum from a component after shutdown.
6. When the draw disappears confirm fix by reconnecting everything except suspected bad part. Then retest to see if draw stays normal.

Common parasitic draw patterns (symptom → likely causes)

Math sanity check: is your measured draw enough to kill the battery in 2–3 days?

In quick-and-dirty terms: time (hours) ≈ usable battery capacity (Ah) ÷ draw (A). A quick estimate can give you a ‘sanity check’ on whether your draw reading matches the symptom you’re observing. Usable capacity is often much lower than the battery’s rated Ah once you factor in cold temperature, battery age, and the fact that many cars won’t crank reliably once the battery drops well below full charge.

Example: If you assume ~30 Ah “usable” before no-start (varies a lot) and you measure a steady 0.3 A draw, then 30 ÷ 0.3 ≈ 100 hours (~4 days). That’s very consistent with ‘dies after a weekend.’
If you measure only 30 mA (0.03 A) and the battery dies in 2 days, then the draw likely isn’t your full story—suspect a weak battery, or a charging issue, and/or an intermittent draw you didn’t catch.

Common mistakes that waste hours (avoid these)

• Not waiting for sleep mode: You’ll chase a draw that’s just normal module activity.
• Opening the doors while testing: Interior lights and modules wake up instantly; use taped switches or latched strikers.
• Meter leads in the wrong jack: This can blow the meter fuse or give nonsense readings.
• Starting the test on the mA range: Always start on the 10A range first; then step down once you’re sure the current is low enough.
• Pulling fuses on a newer car and repeatedly waking modules: Do the fuse voltage-drop method instead if you keep resetting sleep state.
• Fixing the draw but not recharging the battery properly: A partially charged battery can still fail to crank, making it look like the draw remains.

(Mostly after you defeat the draw: how to confirm it’s really gone)

1. Reconnect everything properly and clear any tools/tape used to defeat switches before starting your vehicle.
2. Fully charge the battery as a proper charger is a bit better than hoping it charges by being left to idle.
3. Measure key-off draw again from scratch: set everything up for sleep, wait, then read the fully-asleep value. Compare to service specs if you have them to work from. Otherwise, confirm that draw is at least dramatically way lower by the eyeball method, and in the normal-looking range.
4. Do a real-world validation: park it for the same 2-3 day window that used to kill it, then come back expecting it to start normally if you have it all set right.

(When to stop yourself and make the call for the pro)

• You have a random draw that doesn’t do something like the wipers come on sometimes before you did some damage.
• Time to get serious logging equipment and learn more.
• The draining fuse comes from the BCM and powers like a dozen module bits. Unplugging components causes a slew of new faults.
• The vehicle is still under warranty or has advanced driver-assist systems where a faulty electrical fix can cause steep prices to fix.
• You don’t want to crank the meter.
• You don’t want sparks or burn your multimeter out.
• You don’t know how to use that sucker where you measure current in series at the battery voltage.

FAQ

References

1. Fluke — How to Find Parasitic Battery Drain with a Multimeter
2. ALLDATAdiy — Parasitic Draw Test (including voltage-drop across fuses and sleep time considerations)
3. NOCO — Parasitic Battery Drain (example normal draw ranges and general guidance)
4. Hagerty — Circling the Drain: Sussing out parasitic drains requires a multimeter and creative sleuthing
5. MOTOR Magazine — Trouble Shooter: Parasitic Battery Draw (voltage drop across fuses concept)
6. Continental Battery — The impact of parasitic draw on your car battery’s lifespan