The impact of El Niño–southern oscillation (ENSO) on Indian foodgrain production was analysed for the period 1950–99. The inverse relationship between sea-surface temperature (SST) anomalies from June to August (JJA) over the NINO3 sector of the tropical Pacific Ocean and Indian foodgrain production anomalies (r = −0.50) was significant at the 1% level. During the warm ENSO phase, the total foodgrain production frequently decreased (12 out of 13 years) by 1 to 15%. In 10 out of 13 cold ENSO-phase years, the total foodgrain production increased from normal. The relationship between the SST-based NINO3 ENSO index and the Kharif season (June–September) foodgrain production anomalies (r = −0.52) was greater than for the Rabi season (October–February) foodgrain production (r = −0.27). The ENSO impact on rice production was greatest among the individual crops. The average drop in rice (Kharif season crop) production during a warm ENSO-phase year was 3.4 million tonnes (7%). In a cold ENSO-phase year the average production increase was 1.3 million tones (3%). Wheat (Rabi season crop) production was also influenced by ENSO, as it depends on the carryover soil water storage from the Kharif season. Sorghum and chickpea production are not significantly influenced by ENSO extremes. Inter-annual fluctuation of the gross value of Indian foodgrain production was very large, reducing up to US$2183 million in a warm ENSO year and increasing up to US$1251 million in a cold ENSO year. On average, a warm ENSO year costs US$773 million and a cold ENSO year had a financial gain of US$437 million from normal. The cumulative probability distributions of foodgrain production anomalies during cold and warm ENSO phases are shifted positively or negatively, relative to the neutral distribution. The warm ENSO forcing significantly (1% level) reduced the probability of above-average production. The cold ENSO forcing moderately increased the above-average foodgrain production over the neutral ENSO phase (5% level). A simple conditional probability forecast based on annual and JJA NINO3 SST predicted the category of foodgrain production in 11 of the 14 years. The results demonstrated that the relationship between NINO3 ENSO index and foodgrain production could be used for agricultural applications and policy decisions on food security for the rapidly growing population in India.