During the northern summer the Tibetan Plateau is a heat source for the atmosphere, and the Equatorial Pacific Ocean Cold Tongue is a heat sink, both contributing to the thermal forcing of large-scale quasi-zonal atmospheric circulation. For the period 1954–1979 interannual variability of Indian monsoon rainfall (IM) is found to correlate highly with the thermal contrast between the Tibetan Plateau (TP) and the Equatorial Cold Tongue (ECT). An index of this thermal contrast (SLO for sea-land oscillation) is defined as the departure from the mean of the difference between the ground surface temperature of the Tibetan Plateau (Ts,) and the surface temperature of the Equatorial Cold Tongue (Tw). The interannual variability of Ts and Tw are of comparable magnitude and statistically independent of each other. The index of contrast (SLO) that contains them both correlates more strongly with Indian monsoon rainfall (0.61) than does either Ts(0.30) or Tw(0.52). The Ts and Tw each contributes about equall... Abstract During the northern summer the Tibetan Plateau is a heat source for the atmosphere, and the Equatorial Pacific Ocean Cold Tongue is a heat sink, both contributing to the thermal forcing of large-scale quasi-zonal atmospheric circulation. For the period 1954–1979 interannual variability of Indian monsoon rainfall (IM) is found to correlate highly with the thermal contrast between the Tibetan Plateau (TP) and the Equatorial Cold Tongue (ECT). An index of this thermal contrast (SLO for sea-land oscillation) is defined as the departure from the mean of the difference between the ground surface temperature of the Tibetan Plateau (Ts,) and the surface temperature of the Equatorial Cold Tongue (Tw). The interannual variability of Ts and Tw are of comparable magnitude and statistically independent of each other. The index of contrast (SLO) that contains them both correlates more strongly with Indian monsoon rainfall (0.61) than does either Ts(0.30) or Tw(0.52). The Ts and Tw each contributes about equall...