Compression testing of open cellular carbon foams is a challenge. The compression load between the foam specimen and the loading platen applies high stress concentration at the contact region and causing premature failure. Thus, the measured compression modulus and strength tend to be much lower than the real values of the foam. This paper presents a simple method to avoid or mitigate the stress concentration at the contact between the cell walls and the loading platen by end-potting of samples. The end-potting involves coating the top and bottom contact surfaces of the specimen by a thin layer of resin followed by curing. This offers a smooth contact between the specimen and two loading platens by potting the cell walls in resin thus reducing the contact stresses on the wall. Both the end-potted and standard (non-potted) specimens of low-density (CFoam20 and CFoam20G) and high-density (CFoam30 and CFoam30G) carbon foams were tested in compression. The standard specimens failed by crushing of edges of cell walls followed by progressive crushing of cells. The end-potting prevented local crushing of cells and failed suddenly in a shear failure mode like a brittle material. Compared to standard specimens, end-potted specimens resulted in higher modulus and strength of foams, produced consistent and reproducible test data. Thus, end-potting is recommended for compression testing of open-cell carbon foams and similar brittle foams.