In recent years, underground mines in Nevada are increasingly exploiting in weak mineralized zones at greater depths that are intensely fractured and highly altered. The mechanical behavior of these rock masses ranges between weak rock and very stiff soil. A common limitation for design of underground mining excavations in these types of rock masses is absence of in situ geotechnical data. This limitation is generally overcome by estimating in situ mechanical behavior from empirical relationships so that the continuum-based numerical methods can be used to evaluate ground support designs. Because of the cost, time, and specialized equipment involved, historically in situ tests have not been performed in these underground mines. Predictive rock mass modulus relationships that are currently available in the literature are derived from field testing of predominantly goodquality rock masses. Consequently, there is limited confidence in using these models for rock masses with Rock Mass Ratings less than 45. In order to overcome some of these limitations, a portable plate loading device (PPLD) was designed and fabricated. The PPLD allows one to perform low cost and relatively quick in situ deformability tests to be performed on weak rock masses in underground mines. Test procedures and data reduction methods were developed to limit potential sources of error associated with the PPLD test. A total of fourteen plate loading tests were performed in weak rock masses at two different active underground mines in Nevada, USA. The resulting the test data were compared to eight published empirical rock mass modulus relationships to determine which, if any, of these relationships are sufficiently accurate for estimating modulus in similar geotechnical conditions. Only two of these relationships were found to be sufficient for first-order estimations of in situ modulus.