Measuring Resistance Is One of Those Things That Sounds Trivial Until You Get Weird Readings
I spent three weeks last year debugging a PCB where every component tested fine individually but the board wouldn't power on. The issue ended up being a 0.47 ohm shunt resistor that looked perfectly normal on the schematic but had cracked internally. It was showing 4.2 ohms instead of 0.47 because the solder joint on one leg had cold-cracked months ago. This is why checking resistance properly matters more than people think. Start by making sure the circuit is completely dead. I mean actually dead. Turn off power, wait at least thirty seconds for capacitors to discharge, and then verify with the meter itself that there's no voltage present. Measuring resistance on a live circuit will blow the fuse in your meter or worse, damage the component you're testing. I've replaced two multimeter fuses doing exactly this before coffee. Never again. Set your multimeter to the resistance setting, usually marked with the omega symbol. If your meter has auto-ranging, just turn the dial to ohms. If it's manual ranging, start on the lowest range—200 ohms is a good default—and work up from there if you get an overload reading. A reading of "1" or "OL" on the display means the resistance is higher than your current range can measure, so switch to a higher setting.
Here's the part most guides skip: you need to isolate the component. Desolder at least one leg of the resistor before measuring, or you're measuring the parallel combination of that resistor with everything else connected to it. I once measured what I thought was a 10k ohm pull-up resistor and got 3.2k. Turned out the microcontroller pin tied to the same net had about 6.8k of input impedance in parallel, throwing off the reading entirely. If you don't lift a leg, your reading is meaningless for that component. Touch the probes firmly to each end of the component. For through-hole resistors, you can often test them in-circuit on the component side without lifting legs if you're careful about accounting for parallel paths. Surface mount parts are a different story—get good hook probes or fine-point tips, and take your time. A loose connection between probe and pad will give you bouncing readings that look like a failing component when it's just bad contact. Write down your readings. Not in your head, not "somewhere on the bench." I use a quick notebook with date, component designator, and expected versus measured values. This habit saved a project when I found three resistors out of tolerance on a batch I'd bought from a questionable supplier. All were within 5% of their stated value, which should have been the first red flag for components rated at 1% tolerance.
The delta mode on better meters is worth mentioning if you have one. It subtracts the probe resistance from your reading automatically. Without it, a cheap pair of test leads might add 0.3 ohms to every measurement below 200 ohms, which matters when you're checking shunt resistors or low-value current-sensing networks. Just touch the probes together, note the reading, and subtract it manually if your meter doesn't handle it.
Get the Full Details
What Actually Goes Wrong When You Measure Resistance
Temperature affects resistance. Most resistors have a temperature coefficient, typically 50 to 200 ppm per degree Celsius. If you're working with precision circuits or measuring components that have been sitting in sunlight versus shade, the difference can be noticeable. A 10k resistor with a 100 ppm/°C coefficient will shift by 0.1% for every degree of temperature change. That's not dramatic for general work but it matters when you're trimming reference voltages or building measurement equipment. Lead inductance and capacitance become relevant at higher frequencies, but for DC resistance measurement, the main concern is thermoelectric voltage. When you connect dissimilar metals—copper probes to a nickel-plated resistor lead—you generate a small Seebeck voltage that can throw off low-ohm measurements. This is mostly a problem below 1 ohm. Four-wire Kelvin sensing eliminates this entirely, but most hobbyists and even some technicians don't have that capability on their meters. Here's a practical edge case I ran into recently: measuring the resistance of a heated component. I was testing a power resistor on a motor controller board and the reading kept drifting downward as the resistor warmed up. The component was a 2 ohm current-sensing resistor with a negative temperature coefficient. Every time I took a measurement, the meter's test current was heating it slightly, changing the resistance, and I was chasing my own tail for about ten minutes before I realized what was happening. The fix was using a meter with a lower test current or taking the reading quickly and moving on.
Fused resistors are another trap. A resistor that shows infinite resistance on an ohmmeter might not actually be open. Carbon composition resistors in particular can develop high-resistance faults that look like partial opens. The resistor still conducts current but at a significantly reduced level. I found one of these in a vintage amplifier restoration project that was reading 47k when it should have been 4.7k. The component was physically intact, no discoloration, no cracking. Just degraded internally from age and heat cycling over forty years.
Common Mistakes That Waste Time
Forcing the probes onto hot components is probably the biggest amateur mistake. The heat can melt insulation, damage the probe tips, and give you a dangerously wrong reading because the component temperature has shifted during measurement. Let circuits cool before testing resistance whenever possible. If you can't wait, use insulated probes and minimize contact time. Assuming the meter is calibrated is another one. I checked my meter against a known 10k standard resistor last month and it was reading 10.12k. That's a 1.2% error, which seems small until you're working on something that requires accuracy. Digital multimeters drift over time, especially budget models. If your readings seem consistently off, compare against a known good reference and factor the offset into your work, or send the meter out for calibration. Don't ignore the meter's own specifications. A $30 multimeter might claim 0.5% accuracy on the 200 ohm range, but that number comes with caveats about resolution and noise. On cheaper meters, the last digit will jump around unpredictably on high-resistance measurements. If you're trying to measure something in the megaohm range and the reading bounces between 1.2M and 1.8M, that's not the component. That's your meter struggling with input impedance and environmental noise. Shield the leads, keep them short, and maybe just accept that you won't get a precise reading at that level with that equipment.
Also, don't touch the probe tips together while measuring high resistance. Your body is a conductor, and touching both tips simultaneously puts your body resistance in parallel with whatever you're measuring. For anything above 100k ohms, this effect becomes significant. Hold the probes by the insulated shafts, not the metal tips. I know it feels awkward at first but it makes a real difference in the readings you get.
When Resistance Measurement Isn't Enough
Sometimes an ohmmeter reading tells you nothing useful about whether a component will actually work in a circuit. A resistor can measure perfectly within tolerance and still fail under load due to power rating issues, thermal stress, or parasitic inductance. And vice versa—a resistor reading slightly off value might work fine in its intended application where that deviation doesn't matter. For diodes and transistors, resistance measurements are even less reliable. These are nonlinear devices, and the resistance you measure depends heavily on the meter's test voltage and current. Different meters use different test conditions, which is why comparing resistance readings across two different multimeters for the same diode can give wildly different results. Use the diode test mode for semiconductors instead, which applies a known forward voltage and measures the resulting current. When you need more than just resistance values, consider tools like an LCR meter for inductance and capacitance, or a component analyzer that characterizes semiconductors directly. But for everyday troubleshooting and verification, a decent multimeter on the resistance setting will handle the vast majority of cases. Just remember that the number on the display is only as good as your technique and your understanding of what's actually connected to the component you're probing.
The bottom line is that checking resistance is straightforward in theory and occasionally frustrating in practice. Get the circuit de-energized, isolate the component, account for parallel paths, and trust but verify your meter's accuracy. Everything else is just troubleshooting what goes wrong when those basic steps aren't followed carefully enough.
