A plasmid is a small DNA molecule that is physically separate from, and can replicate independently of, chromosomal DNA within a cell.
In general, in eukaryotes, episomes are closed circular DNA molecules that are replicated in the nucleus. Viruses are the most common examples of this, such as herpesviruses, adenoviruses, and polyomaviruses. Episomes in eukaryotes behave similarly to plasmids in prokaryotes in that the DNA is stably maintained and replicated with the host cell.
(Reposting my comment as an answer since it seems to be what was required.)
A DNA molecule that replicates independently of chromosomal DNA is an episome. By this definition a plasmid is (usually) an episome. If a plasmid integrates into a chromosome by some mechanism (as for example in Hfr strains of E. coli where the F plasmid is integrated) the plasmid loses its episomal status.
The term "episome" was first coined by Joshua Lederberg to describe the nature of the F element in E. coli.
I'll have to disagree with Alan. If the plasmid is integrated into a chromosome it doesn't lose its episomal status. Actually the integration of the plasmid is a characteristic of an episome.
This is a quote from this article: http://www.qub.ac.uk/mlpage/courses/level3/plasmidhistoryreview.pdf
"... For a decade or more it was confused with ‘‘episome,’’ although that was carefully crafted (F. Jacob and E. L. Wollman, 1958, C. R. Acad. Sci. 247, 154–156) to mean agents with traffic in and out of chromosomes.
Also this book (https://books.google.pt/books?id=YnXB54RtrR8C&pg=PA132&dq=episome&hl=pt-PT&sa=X&ei=oKZgVd3HEZOu7AaeoYPABQ&ved=0CCoQ6AEwAg#v=snippet&q=episome&f=false) clearly mentions that an episome is a genetic element able to exist in an integrated form (associated with the chromosome) but also with it's own origin of replication (autonomous form). So in fact an episome can be a plasmid or a viral genome, but not all genetic elements have this ability.
