Power systems have been subjected to several changes due to a constant evolution over the last years. The changes that come along with this evolution include several aspects such as the introduction of variable energy generation from renewable sources, changes in the network's topology, storage, new regulation policies, liberalization of markets, constant trading of electrical energy between countries among others. These changes, combined with a context of growing demand and aging network infrastructure can result in more stressed power systems, uncertainty and potentially more risks that could lead to blackouts. These new conditions have to be taken into account when assessing the reliability of the transmission system in order to have a complete insight of what is happening in the network and to identify potential causes of problems. This thesis provides a methodology using state enumeration to assess reliability in the transmission network in the context of long term planning using a probabilistic approach with the aim to identify potential improvements for current practices based on deterministic approaches. This methodology considers three main aspects: i) Failure probabilities of individual components (i.e transmission lines) based on historical data and operational conditions. The failure probability originated from operational conditions is obtained by combining expert's opinion and available data by applying Bayesian data analysis; ii) Contingency selection to obtain a representative list of contingencies to be analyzed. The screening of contingencies for the list is performed based on the failure probabilities and accepting a certain level of residual risk. The residual risk is calculated by assuming a lost load and finding the probability of the contingency that leads to that lost load; iii) Determination of the effect and the associated risk of contingencies. The purpose of this part is to quantify line overloadings in terms of money to obtain the total cost linked to a certain set of contingencies.The monetization of the effect is performed by using a value of lost load (VoLL). This monetization is first performed without the implementation of re-dispatch actions assuming that overload is lost load. Later on, re-dispatch actions for conventional generators are considered, the new cost of re-dispatch plus lost load is obtained. Furthermore, a comparison between the two costs is performed. Re-dispatch actions are implemented by using a linear programming approach with the selection of an objective function for the minimization of the amount of conventional generation. Risk of lost load is obtained in terms of money by the multiplication of the probability of the contingency times the cost. The proposed methodology obtains reliability in terms of system adequacy by investigating line overloads and lost load. Further on, this methodology is tested on a test model for the 380/220kV grid topology of the transmission grid of the Netherlands. The results show the variation of indicators of risk of lost load and guidelines to continue developing probabilistic approaches for reliability analysis. The computational burden and the time employed for this kind of approaches is still a challenge to be tackled.