From October 1, 2007, the new bridges on federal-aid funded projects are mandated to be designed to meet American Association of State Highway and Transportation Officials (AASHTO) Load and Resistance Factor Design (LRFD) Bridge Design Specifications. LRFD is a simplified form of reliability-based design. By multiplying calibrated factors to load and resistance components, the designed structure will maintain a specific level of reliability (or probability of failure). By concept, the load and resistance factors should be calibrated by large number of test data; however, they are often unavailable in geotechnical engineering. Significant efforts are needed to calibrate load and resistance factors based on test data of good quality. In this study, 26 O-Cell test data were collected from Kansas, Colorado, Missouri, Ohio, and Illinois. Seven methods available in the literature were selected to estimate the load capacities of 25 out of 26 drilled shafts. The "FHWA 0.05D" method was found to yield the closest and conservative predictions of the nominal resistances to the representative values; therefore, it was adopted in this study when calibrating the resistance factors for Strength Limit State design. These test data were analyzed and used to calibrate side and base resistance factors for drilled shafts in weak rock. Resistance factors were calibrated at two different target reliability indices: 2.3 (i.e., failure probability, Pf ~ 1/100) for shafts with greater redundancy and 3.0 (Pf ~ 1/1000) for shafts with less redundancy. Side resistance factors were calibrated from two different datasets of measured resistance: total side resistance and layered unit side resistance. The resistance factors calibrated from layered unit side resistance are considered more reliable, therefore, they are recommended for design. The recommended resistance factors from this study are compared with those in AASHTO specifications. Some of those calibrated resistance factors from this study are considerably lower than those in AASHTO specifications. The main reasons for such lower resistance factors are mainly attributed to the low efficiency of the FHWA design method and the limited quality and number of O-Cell test data. These resistance factors may be improved by increasing the size and the quality of the test data in the future. At present, field load tests on drilled shafts are recommended as an alternative to using lower resistance factors, which will also accumulate more test data for future improvement.