Counting statistics investigates the probability P(n,t) that a number n of electrons traverse a nano-scale conductor during a time span t. It is equivalent to consider the zero frequency charge or current correlators, the so-called moments and cumulants, in principle up to infinite order. In this thesis we investigate several aspects of electronic correlations due to interactions. First we investigate the influence of interactions on the counting statistics, considering a generic two-terminal conductor. We show that if the factorial cumulants oscillate as functions of any system parameter or time, then the electrons must be interacting. This statement may be verified in Coulomb blockaded quantum dots, where it is possible to monitor the traversal of electrons in real-time. Moreover, we use a Markovian master equation to describe the first experiment on counting statistics of Andreev events, where two electrons tunnel accross a tunnel barrier between a superconducting lead and a normal metallic island. The statistics are strongly super-Poissonian, reflecting that Andreev events occur in avalanches of different sizes. Finally, we consider finite frequency current noise and show that the noise spectra are in general asymmetric in the applied bias voltage. Using a higher order fluctuation relation, which is an extension of the fluctuation dissipation relation to the non-equilibrium transport regime, we show that this asymmetry is due to a broken electron-hole symmetry, resulting in a finite rectification. We point out that this can occur either due to an asymmetrically applied bias, but more importantly, due to interactions and an inherent chirality of the conductor.