A time-dependent primitive equation model with specified zonal flow, mountains, and diabatic heating is described. The model is used to investigate the effects of mountains, diabatic heating, and the nonlinear interactions of their responses in the June–August large-scale global circulation. Particular emphasis is placed on the Asian summer monsoon, a dominant feature of the global circulation at this time. The prescribed heating is calculated as a residual in the time-mean thermodynamic equation from ECMWF data, and the model employs a linear surface drag that is enhanced over the land. The integration is initiated with an observed June–August zonal-mean flow that is maintained throughout the integration. A smoothed earth topography is raised from days 0 to 5 with hydrostatic adjustments made to temperature and surface pressure. When the integration is continued in the absence of diabatic heating, an almost steady-state flow pattern exists until about day 20 when instability begins to dominate. If, however, the diabatic heating is turned on at day 5, the direct response to the heating evolves over the next few days. Many features of the June–August global circulation are reproduced, and timescales for the atmospheric response to imposed heating are seen in the model results. With transients playing little role in the model simulation, the realism of the modeled circulation implies that resolved transients are not of vital importance to the large-scale time-mean global circulation during this season. The realism also implies that parameterizations within the model are sufficiently accurate for the investigation. A linear version of the model, forced with diabatic heating but with no mountains, is found to reproduce the upper-tropospheric circulation. This result is significantly different from the December–February circulation where previous authors have highlighted the importance of both orography and nonlinearity. The subtropical highs of the lower-tropospheric circulation are also given by this linear heating-only model, but their definition and amplitudes are improved when mountains and the nonlinear interactions of the orographic and thermal responses are included. These interactions are found to be essential if the low-level monsoon inflow is to be simulated realistically.