Large Scale Structure: Theories of the origin and the evolution of the universe predict the statistical properties of the dark matter distribution on the largest scales. So if we could measure the dark matter clustering we could learn about the Universe composition, its evolution and even about inflation. In particular the shape of the power spectrum contains information about cosmological parameters such as matter density , baryon density, n,
dn/d lnk. Higher-order correlations can be used to test the nature of the initial conditions such as non-gaussianity. While we cannot easily observe the dark matter distribution directly and the galaxy distribution could be a biased tracer of the dark matter distribution, higher-order correlations can be used to measure this bias. I have developed the bispectrum technique and I have applied it to the 2 degree field (2dF) galaxy redshift survey to constrain the bias parameter: 2dF galaxies are faithful tracers of the dark matter distribution. I have used galaxy surveys data (2dF and SDSS) in combination with other data sets to constrain cosmological parameters. I have also investigated the potential of large scale structures to constrain the nature of the initial conditions.