If you have ever taken your baby's temperature with a forehead thermometer, gotten 99.8°F, then taken it rectally and gotten 100.6°F, you are not doing anything wrong. Different methods measure at different body sites, and each site runs at a slightly different temperature.
Rectal temperature is the closest approximation of core body temperature — the temperature inside your baby's body where the organs are. It is the most accurate and the reference standard against which all other methods are compared.
Temporal artery (forehead) thermometers measure the blood flowing through the temporal artery just beneath the skin. They are fast and non-invasive, which is why parents love them. They are reasonably accurate for babies over 3 months, but they can be affected by sweat, ambient temperature, and technique. If your baby has a sweaty forehead, the reading may be falsely low.
Axillary (armpit) temperatures are consistently lower than rectal — typically by 1 to 2 degrees Fahrenheit, though the exact difference varies. This is because the armpit is an external measurement point and is affected by ambient temperature and how tightly the arm is held against the body. Armpit readings are useful as a quick screen ("does this baby feel warm to me?"), but they are not reliable enough to make clinical decisions, especially in young infants.
Tympanic (ear) thermometers measure infrared heat radiating from the eardrum. They can be accurate in older children and adults, but they are not recommended for babies under 6 months. An infant's ear canal is small and curved, making it difficult to aim the sensor correctly. Earwax can also interfere with readings.
The takeaway: for babies under 3 months, use rectal. For babies over 3 months, rectal or temporal artery are both acceptable. For a deeper look at technique, see our step-by-step guide on how to take a baby's temperature.