Because the diurnal cycle is a fundamental, forced mode of variability in the atmosphere, it has been suggested as an attractive test case for climate models. In the central Pacific, away from the influence of landmasses, observations show small but significant diurnal variation in many quantities, including wind and convection. Precipitation in the ITCZ reaches its maximum in the early morning hours, lagged by high-level cloud cover by a couple of hours. Surface winds show more complicated diurnal phasing: the zonal component is dominated by the semidiurnal tide with maxima in the early morning and afternoon while the meridional component is dominated by the diurnal harmonic that maximizes mostly during the daytime, but with some variation across the basin. This study shows that the Community Atmosphere Model (version 5) captures these observed features with enough verisimilitude to motivate better understanding the mechanisms of the diurnal cycle of wind and convection at work in the model. We make use of a series of 5-day hindcasts, i.e. the CAPT framework. The hindcasts initialize a realistic large-scale circulation that evolves slowly toward the model's climatology. Within this relatively realistic large-scale environment, the fast physics processes (e.g., clouds and turbulence) rapidly influence the thermodynamic state; previous studies have shown that a model's hindcast errors are related to its climate errors. Over the central Pacific, initial errors appear to saturate in about a day, so analysis is performed over the subsequent few days of the hindcasts. Temperature and moisture budgets are examined to establish the connection between the low-level inflow from the trade-wind regions and the convection of the ITCZ. Sensitivity experiments repeat the hindcasts with physics changes that help elucidate the relative roles of upper and lower level clouds and convection in the diurnal variation of deep convection. Results suggest that the diurnal cycle of convection over tropical oceans provides a view into, and perhaps a test for, interactions between parameterized physics and the large-scale circulation.