A paired t-test is a statistical method used to compare two related groups to determine if their means differ significantly. It is commonly used when the same subjects are measured under two different conditions or at two different times, accounting for the natural pairing of observations.