El Nino-Southern Oscillation (ENSO) consists of irregular episodes of warm El Nino and cold La Nina conditions in the tropical Pacific Ocean(1), with significant global socio-economic and environmental impacts(1). Nevertheless, forecasting ENSO at lead times longer than a few months remains a challenge(2,3). Like the Pacific Ocean, the Indian Ocean also shows interannual climate fluctuations, which are known as the Indian Ocean Dipole(4,5). Positive phases of the Indian Ocean Dipole tend to co-occur with El Nino, and negative phases with La Nina(6-9). Here we show using a simple forecast model that in addition to this link, a negative phase of the Indian Ocean Dipole anomaly is an efficient predictor of El Nino 14 months before its peak, and similarly, a positive phase in the Indian Ocean Dipole often precedes La Nina. Observations and model analyses suggest that the Indian Ocean Dipole modulates the strength of the Walker circulation in autumn. The quick demise of the Indian Ocean Dipole anomaly in November-December then induces a sudden collapse of anomalous zonal winds over the Pacific Ocean, which leads to the development of El Nino/La Nina. Our study suggests that improvements in the observing system in the Indian Ocean region and better simulations of its interannual climate variability will benefit ENSO forecasts.