The validity of a test, or the degree to which a test actually measures what it is initially intended to measure (Hughes, 1989: 22; Brown, 2001: 387), is a complex criterion in the field of testing. With the ever growing list of recognised validities in the academic field of language testing, Owen et al., (1997: 20) highlights that this list is often dispiriting due to the abundance of choice and approved validities. Amongst this selection, the most pertinent types of validity for the purpose of this paper are ‘construct’, ‘criterion-related’ (concurrent and 7 predictive), ‘content’ and ‘face’ validities. The phenomenon of backwash will also be discussed.