A single resistance factor is used in typical steel design recommended by the American Institute of Steel Construction. In most cases, the steel shapes are homogeneous and isotropic. Hence, a single resistance factor is sufficient. However, in the case of composite sections, such as concrete-encased or concrete-filled steel beam-columns, a single resistance factor for these beam-column designs may not be adequate. Studies have been shown that variations in steel strength and concrete strength are different, and so are their respective probability distributions. As the number and height of high-rise buildings continue to increase, so will the use of steel-concrete composite structural members. Thus, it is appropriate to examine the adequacy of a single resistance factor. In this paper, the focus is on beam-columns. The reliability index based on partial resistance factors for steel and concrete strengths was compared with that based on a single resistance factor. In addition, the study also considered the effect of correlation between axial load and bending moment due to eccentricity. Using the target reliability indices determined from current overall resistance factor design, a set of partial resistance factors was developed for each category of composite beam-column sections. The resistance factor associated with the concrete portion of the composite section is consistent with the resistance factor of reinforced concrete compression member given in ACI 318-02. While the overall resistance factor applied to column strength does make the beam-column design simpler, the slightly more complex partial resistance factor approach presented could lead to more efficient designs.