Weather extremes are expected to increase as climate warming continues, yet how productivity consequences may differ for agricultural crops and semi-natural vegetation in heterogeneous agricultural landscapes remains uncertain. This study evaluated aboveground biomass productivity of dominant crops (corn, soybean, potato) and semi-natural vegetation (wetlands, grasslands) during an extremely dry vs. wet growing season in two Wisconsin (USA) agricultural landscapes. We asked: (1) How does sensitivity to weather differ extremes among crops, wetlands and grasslands? (2) How does sensitivity vary between landscapes dominated by loamy vs. sandy soil? (3) To what extent does irrigation mitigate the sensitivity of croplands to drought? Using the Normalized Difference Vegetation Index (NDVI) derived from Landsat imagery as an indicator of aboveground biomass productivity, we compared an extremely dry (2012) and unusually wet (2018) growing season in Wisconsin (USA). We randomly sampled mid-summer (July 15-31) NDVI during 2003-2021 for each vegetation type in the Yahara Watershed (loamy soil) and Central Sands (sandy soil) to establish a long-term baseline. Sample locations were consistent for the two focal years. We also computed NDVI anomalies relative to long-term NDVI in both years and compared frequency distributions of NDVI between rainfed and potentially irrigated croplands in the Central Sands. Differences between NDVI distributions in the dry and wet year were most pronounced for corn, potato and semi-natural grasslands (all D > 0.5) and least pronounced for soybeans and wetlands (all D < 0.3). During the dry year, NDVI anomalies from the long-term median were also greatest for corn, potato and grasslands (-17%, -13% and -22% respectively). Additionally, vegetation was most sensitive to drought in the Central Sands vs. Yahara landscapes (average of 16% vs. 6% reduction across vegetation types, respectively) and insensitive to wetter conditions in both landscapes (anomalies < 1%). Irrigation, which is more prevalent in the Central Sands, mitigated but did not eliminate NDVI declines in all crop-growing areas in the dry year. Weather extremes influence agricultural crops and semi-natural vegetation types differentially, and NDVI indicated lower biomass and thus likely reduced crop yield during the dry year. Vegetation responses also differed between regions, indicating the importance of accounting for heterogeneity in crops, seminatural vegetation and regional context when assessing productivity responses to extreme weather.