Characteristic of a number stored in twice the amount (two words; typically 8 bytes) of computer memory that is required for storing a less precise (single-precision) number. Commonly handled by a computer in floating-point form.