One of the most intriguing recent results in physics is the growing evidence that an unknown energy field and an unknown kind of matter are the major components of the Universe (70% and 30%, respectively; see e.g. Riess et al. 1998, Spergel et al. 2007). Understanding and estimating the precise value of the energy density of the two above components, together with measuring the Hubble constant that determines the age of the Universe, is a major goal of modern astrophysics.

An interesting method for estimating these parameters is strong gravitational lensing of quasars (QSOs). As shown by Refsdal (1964), H0, !m and !! can be measured based on the time delay ("t) between multiply lensed images of QSOs, because "t depends on H0 and on the distances to lens and source, hence!m and !!. Determination of cosmological parameters using gravitational lensing suffers from some degeneracies, but it is based on well understood physics and unlike distance ladder methods there are no calibration issues. Moreover, it has an advantage over some of the leading methods (such as Type Ia SNe) in that it is a purely cosmological approach.

In this thesis, the property of strong gravitational lensing - time delay - is extensively explored.
Strong gravitational lensing, and in particular time delays, are investigated here both theoretically and observationally. The focus is on the time delay as a cosmological and astrophysical tool to constrain H0, !m, !!, and to measure the masses of lensing galaxies.
The first Chapter presents a historical background of gravitational lensing. It explains the process involved in creating the theory of gravitational lenses. It shows how Newton and then Einstein developed the concept and how Refsdal and others made it a cosmological tool. We present in more detail how gravitational lensing influenced the history of physics by being the
first proof of Einstein's theory of gravity. The Chapter ends by discussing the first observational discovery of a strong gravitational lens.