Climate change is one of the most important challenges facing agricultural production, particularly in arid and semi-arid regions, where water scarcity, rising temperatures, recurrent droughts, and hydrological fluctuations can simultaneously increase production risks. The Sistan Plain in southeastern Iran is highly vulnerable to these changes because its agricultural system depends heavily on limited water resources and the variable flow of the Helmand River. This study aims to estimate and distinguish the short-run and long-run effects of climatic variables and water availability on agricultural production in the Sistan Plain. To this end, annual data for six major groups of irrigated crops, including cereals, legumes, industrial crops, vegetables, melons and squashes, and forage crops, were used over the period 1370–1403 in the Iranian calendar (1991–2024). The dataset comprises 204 observations and constitutes a strongly balanced panel with six cross-sectional units observed over 34 years. The variables include crop production, harvested area, annual precipitation, mean annual temperature, the Emberger climatic index, and Helmand River water flow. Given the potential for cross-sectional dependence arising from common climatic shocks and regional environmental conditions, second-generation panel unit-root tests and panel cointegration tests that account for cross-sectional heterogeneity and dependence were employed. The results indicate that the variables are a mixture of I(0) and I(1) processes, with none integrated of order two, I(2), thereby supporting the application of the panel Autoregressive Distributed Lag (ARDL) framework. The Dynamic Fixed Effects (DFE), Mean Group (MG), and Pooled Mean Group (PMG) models were estimated, and the Hausman test identified the PMG specification as the preferred estimator. The long-run results indicate that harvested area has a positive and statistically significant effect on agricultural production, with a 1% increase in harvested area associated with an approximately 0.93% increase in production. Precipitation also has a positive and statistically significant effect at the 10% level, with a coefficient of 1.510. In contrast, mean annual temperature has a large and statistically significant negative effect on production, with a coefficient of −5.308. The Emberger climatic index also has a negative coefficient of −1.372 and is statistically significant at the 10% level, indicating the adverse effect of increasingly hot and dry climatic conditions on agricultural production. The coefficient of Helmand River water flow is −0.033 and statistically significant at the 10% level. However, this finding does not necessarily imply a direct negative effect of increased water flow on production; rather, it may reflect the timing and variability of river flows, differences between recorded river flows and water effectively available to farmers, and constraints in water distribution and utilization. In the short run, changes in harvested area have a positive effect on production, whereas the effects of changes in precipitation, temperature, and the Emberger index are not statistically significant. Moreover, the error-correction coefficient is −0.57 and statistically significant, indicating that approximately 57% of short-run deviations from the long-run equilibrium are corrected within one period. The findings demonstrate that rising temperatures and increasingly unfavorable climatic conditions are among the major long-run threats to agricultural production in the Sistan Plain. Therefore, improving water-use efficiency and irrigation systems, strengthening water storage and groundwater recharge, adjusting cropping patterns, developing drought, heat, and salinity-tolerant crop varieties, and strengthening mechanisms for monitoring and managing Helmand River water resources should be placed at the center of climate-change adaptation policies in the region.