(aka quantum) The amount of run time allocated to a process by the process scheduler. After using up its time slice a process is preempted until it is its turn again. Processes don't have to use up their time slice but may block or yield instead. In some operating systems (OS) using up the time slice reduces the priority of a process.
Small time slices decrease response time, which is good, but increase overhead due to the frequent process switching. Therefore some OSs (e.g. Solaris) increase the time slice of long running processes if they are on their own (i.e. enough time left from interactive jobs).