A multimeter on a car battery is one of the few diagnostics anybody can do at home, and one of the easiest to misread. The number on the screen is a snapshot of state of charge, not of health, and those are different things. Here is what each reading actually tells you, and where the reading stops being useful.
The numbers, with the engine off
Take these at least an hour after the car was last run, or overnight. A battery that has just been charged carries a surface charge that reads high and disappears under load.
- 12.6 volts or above. Fully charged. Nothing to do.
- 12.4 volts. About seventy-five percent. Normal after a few days parked, worth investigating if the car was driven yesterday.
- 12.2 volts. About half. The car will probably still start in mild weather and probably will not at fifteen degrees.
- 12.0 volts. About a quarter, and this is where damage begins. Lead sulfate starts crystallising on the plates and some of that capacity does not come back.
- Below 11.8 volts. Flat. Jump it if you must, but understand you are moving the problem rather than fixing it.
Notice how narrow that range is. The difference between a full battery and a flat one is under a volt, which is why a reading of 12.2 sounds fine to most people and is not.
The numbers with the engine running, which measure something else entirely
Start the car and the meter is no longer reading the battery. It is reading the alternator.
- 13.7 to 14.7 volts. Normal charging. Modern cars vary within this deliberately, so a figure that moves around is not a fault.
- Under 13.0 volts. The alternator is not keeping up. The car is running off the battery and will eventually stop.
- Over 15 volts. Overcharging, usually a voltage regulator. This boils the electrolyte out of the battery and cooks anything electronic downstream of it.
One modern complication worth knowing: many newer vehicles use a variable or smart charging strategy that deliberately drops charging voltage when the battery is full, to save fuel. On those, 12.9 volts at idle is not necessarily a fault, and reading it as one is how people end up buying an alternator they did not need.
Where voltage stops telling you the truth
This is the part most guides leave out. Voltage measures charge, not capacity. A battery with half its plates sulfated can sit at a perfect 12.6 volts with nothing drawing on it, and then collapse to 9 volts the moment the starter turns. It will read as healthy right up until it strands you.
What actually measures health is a load test or a conductance test: putting a known demand on the battery and watching what the voltage does. The number that matters is how far it sags while cranking. Below about 9.6 volts during a crank at normal temperature, the battery is done regardless of what it read at rest.
That is a two-minute test with the right tool and it is not something a multimeter can do. It is also free here, and worth having done in October rather than January.
A parasitic drain looks like a bad battery and is not
If a battery reads fine after charging, tests fine under load, and is still flat after three days parked, the battery is not the problem. Something is drawing current with the car asleep.
Every modern car has a normal quiescent draw for the modules that stay awake, generally somewhere under 50 milliamps once everything has gone to sleep, which can take the better part of an hour after you lock it. A draw meaningfully above that is a fault: a glovebox or trunk lamp that never switches off, an aftermarket accessory wired to a permanent live, a module that will not sleep, or a failing relay.
Finding it means putting a meter in series with the battery, waiting for the car to settle, and then pulling fuses one at a time until the number drops. It is a patient job rather than a difficult one, and it is far cheaper than the third battery.
Michigan makes all of this worse, in two directions
Cold gets the blame, and it deserves some: a battery at zero degrees has meaningfully less available current than the same battery at eighty, while the engine it has to turn is harder to turn. But the damage usually happened in July. Heat is what actually degrades the plates and evaporates electrolyte, and a summer of short trips around Royal Oak that never fully recharge the battery leaves it entering winter with less capacity than the label says. The cold does not cause the failure. It exposes it.
The short-trip problem is real and worth naming. Starting the car takes a large slug of charge, and putting it back takes sustained running. A pattern of five-minute trips means the battery never returns to full, and a battery held at partial charge sulfates. If that is your driving, either take a longer run occasionally or put it on a maintainer.
What to do with all this
If you have a meter, the useful habit is one reading a month with the car cold, written down. The trend matters more than any single number: a battery drifting from 12.6 to 12.4 to 12.2 over three months is telling you something a one-off reading never will.
If it is reading low, charge it properly with a charger rather than by driving around, then test it under load. If it holds, you have a charging or drain problem. If it does not, you have an answer.
And if the starter has started sounding slower than it used to, do not wait for the meter to agree with you. That is the most reliable early symptom there is, and it is usually a few weeks ahead of the morning it does not start at all.