This paper studies how robust Steiner s peak load pricing results are to changes in certain assumptions. Non-storability and periodic demand fluctuations give rise to the peak-load problem. The peak-load problem deals with choosing the optimal pricing scheme leading to optimal output when there is a non-storable good whose demand fluctuates periodically at a uniform price. In the long-run planning point of view the problem also deals with optimal capacity of the system as opposed to short-run where the existing capacity is fixed an thus not subject to determination. I consider the peak-load problem for electric utilities. Steiner s Peak load pricing results involves: (result 1) charging different prices for electricity in different time intervals according to long run marginal cost of generating electricity; (result 2) charge high prices when consumption tends to rise above the level of the capacity and charge lower prices in periods with excess capacity; (result 3) no responsibility for capacity cost imputed to the off-peak customers whose aggregated demand does not press upon capacity; and finally, (result 4) capacity is simply found where equal to peak load. I consider those papers relaxing one of the assumptions of the Steiner model without seriously undermining the insights from the basic model. I will study the implication of relaxing the assumptions: long-run planning point of view, linear costs, capacity fully variable in the long-run, independent demand, a welfare-maximizing social planner, single-technology and periods of equal lengths. I disregard those articles deriving a completely new model where the results are not directly comparable with the results of the Steiner model. I will not study the implication of changing the framework of a static, deterministic partial equilibrium model with exogenously determined demand functions, homogenous agents, no transmission costs, no intra period time varying demand, no storage possibilities, fully divisible capacity and no competitive element. The result 1 of price equal to long run marginal cost is not robust to changes in the assumption of long run planning point of view, linear costs, and fully variable capacity in the long run, dependent demand and the objective of maximizing welfare. When relaxing the assumptions, prices still depend on marginal cost, however not the long-run marginal cost Steiner advocate but the short-run marginal cost. When considering a breakeven welfare-maximizing social planner or a profit-maximizing social planner, prices also depend on elasticity of demand. A breakeven constraint is imposed when there are non-linear costs or fixed capacity in the long run, to ensure the firm at least breaks even. If the monopoly is regulated the prices are also affected by the specific regulation. When dependent demand is considered, prices also depend on the cross-elasticity of demand. However, the relevance for this thesis of altering the assumption of independent demands is questionable. Dependent demand may violate the framework of partial equilibrium model and thus be outside the scope of this paper. With multiple technologies some sort of marginal cost pricing is still relevant. Price is set equal to the marginal cost of expanding the demand in that period. Result 2 of peak price higher than off-peak price follows automatically in the standard model in which there is a welfare objective, the firm has constant returns to scale in production and where capacity is fully variable in the long-run view. In general, nothing as strong as this result can be stated when considering a profit-maximizing monopoly or a breakeven welfare-maximizing social planner. Then prices depend on elasticity of demand and the pricing reversal phenomenon may occur depending on the parameters. Additionally, when we relax the assumption of independent demand, price will also depend on the cross-elasticity of demand and may contribute to pricing reversal. Result 3 of no responsibility for capacity cost imputed to those customers whose demand does not press upon capacity has been criticized on welfare grounds. Off-peak customers are also served by the capacity even if they do not press against the capacity limit. In the single-technology case the result is not valid for the short-run peak load problem, as capacity cost is only related to the long-run peak load problem. The one-technology (i.e. homogenous plant capacity) assumption is crucial for the result that peak users bear all of the capacity costs. When diverse technology is introduced off-peak customers are made to contribute to capacity costs, since they press against the capacity limit to the base-load capacity. Result 4 of optimal capacity found where it is equal to peak demand when optimized is relatively simple, and therefore does survive the different extensions reviewed. For all the extensions of the model, optimal capacity is equal to peak load demand due to the imputed capacity constraint. However, how to find optimal capacity in the extended models is different than Steiner advocate. Optimal capacity is found where the willingness to pay for an additional unit of capacity is equal to the cost of that unit including other components specific for the relaxed assumptions. The robust result of Steiner s peak-load pricing when relaxing the above-mentioned assumptions is to set one price in each pricing period in accordance with the pattern of demand and prices are closely tied to variation in the marginal cost of generating electricity. Optimal capacity is equal to peak load.