Google it, you'll get a much more thorough answer.
Redundant Array of Independant Disks, there are a few different configurations. But the gist of it is you have data copied on multiple disks, so if one fails you have the data (or part of the data) on another disk. The RAID controller does this for you so it is constantly copying the most up-to-date data and files to all drives - this way the CPU isn't bogged down by it. The most common RAID config. are:
0 - (Requires 2 disks) Disk stripping, this provides NO data protection, it just allows your system to run faster because pieces of the data are on each drive, essentially allowing the OS/RAID controller to grab more data at once.
1 - (Requires 2 disks) Mirroring, you have two drives that are identical. If one fails, you replace the failed drive with a new one and rebuild the array based off the other drive that is still good.
5 - (Requires 3 disks) This uses parity to place bits of information on all three drives. As most know all computers are is mathmatical machines, and all data and files are a mixture of 1's and 0's. Parity takes a file apart and creates an equation, the parity of the equation is the total of both halves. So 1+1=2, each number being on a different drive. So if you lose the first 1, we can rebuild the array with the remaining 1 and the parity information of 2.
There are lots of other RAID configs out there, but these are the most common.