Recent studies provide ample evidence of the effectiveness of two-phase spray cooling at dissipating large heat fluxes from electronic devices. However, those same studies point to the difficulty predicting spray performance, given the large number of parameters that influence spray behavior. This paper provides a complete set of models/correlations that are required for designing an optimum spray cooling system. Several coolants (water, FC-72, FC-77, FC-87 and PF-5052) are used to generate a comprehensive spray-cooling database for different nozzles, flow rates, subcoolings, and orientations. High-speed video motion analysis is used to enhance the understanding of droplet formation and impact on the device's surface, especially near the critical heat flux (CHF) point. A previous CHF correlation for normal sprays is modified for both inclination and subcooling effects. A new user-friendly CHF correlation is recommended which shows excellent predictive capability for the entire database. Also discussed in this paper is a new theoretical scheme for assessing the influence of spray overlap on cooling performance.