The major codes that contain recommendations and discussions concerning the design of deep beams, including the ACI Building Code 318-95, the Canadian Code CAN-A23.3- M84, the CEB-FIP Model Code and the CIRIA Guide 2, present design methods that do not cover adequately the dimensioning of this type of beams. Nevertheless, some other codes don’t give any special recommendation. The Brazilian Code (NBR 6118), for instance, just explains that this type of beams should be calculated as a plate in elastic range. Another example is the current British Code BS 8110, which explicitly states that “for design of deep beams, reference should be made to specialist literature”. Because of reasons as those mentioned above, obtaining a rational method not only based on a clear mechanism of failure but taking into account the main parameters that have influence on the ultimate strength of deep beams, has been the purpose of several researchers in the whole world in the last two decades. In this work, some methods for the design of simply supported reinforced concrete deep beams are presented, examined and commented upon. These methods are applied to a total of 37 beams tested in the Laboratory of Structures and Materials (L.E.M) of PUC-Rio, since 1979, and to some beams reported in the literature, in order to yield a method which can predict results of ultimate load closer to those ones obtained experimentally. The future aim is to achieve recommendations that could be proposed to the Brazilian Code. The tests of the 37 beams just referred are included in theoretical-experimental research done by Guimarães (1980), Vasconcelos (1982) and Velasco (1984), which took place in PUC-Rio, orientated by Professor Khosrow Ghavami. Several concluding remarks were obtained in each Master Thesis, apart from one another, but there wasn’t any work that compared these results. The present work is also intended to provide some comparative information regarding the 37 deep beams mentioned above, with the support given by the current literature.