A backup battery provides power to a system when the primary source of power is unavailable. Backup batteries range from small single cells to retain clock time and date in computers, up to large battery room facilities that power uninterruptible power supply systems for large data centers. Small backup batteries may be primary cells; rechargeable backup batteries are kept charged by the prime power supply.

Modern personal computer motherboards have a backup battery to run the clock circuit and retain configuration memory while the system is turned off. This is often called the CMOS battery. The original IBM AT, for example, used a small primary lithium battery to retain the clock and configuration memory. Modern systems use either primary or rechargeable batteries. Primary batteries required periodic replacement; rechargeable types often lasted as long as the system they supported.

Backup batteries are used in uninterruptible power supplies, and provide power to the computers they supply for a variable period after a power failure, usually long enough to at least allow the computer to be shut down gracefully. These batteries are often large sealed lead-acid batteries.

Server-grade disk array controllers often contain onboard cache memory, and provide an option for a "backup battery unit" (BBU) to maintain the contents of this cache after power loss. If this battery is present, disk writes can be considered completed when they reach the cache, thus speeding up I/O throughput by not waiting for the hard drive. This operation mode is called "write-back caching".