Global climate change remains one of the most urgent challenges of the twenty-first century, with its adverse effects disproportionately impacting arid and semi-arid regions where limited resources and persistent water scarcity constrain agricultural development. Multi-criteria evaluation methods that combine the fuzzy analytical hierarchy process (AHP) and the technique for order preference by similarity to ideal solution (TOPSIS) with geospatial techniques have become essential for land suitability assessments that support climate-smart agricultural practices and policy decisions. This study assessed land suitability for rice, millet, and maize by integrating fuzzy-AHP and TOPSIS within a geospatial framework. Ten biophysical and climatic factors were converted into thematic layers in ArcGIS, weighted using fuzzy AHP, and integrated under the FAO land evaluation framework to produce multi-crop suitability maps. Spatial analysis revealed pronounced inter-district variability in agricultural potential. For rice, Manvi had the largest area of highly suitable land (79,312.46 ha), followed by Raichur (53,427.68 ha), while Devadurga had only 13,587.24 ha in this class, indicating greater climatic and soil constraints. Millet suitability was also highest in Manvi (63,412.36 ha), with Raichur (35,127.49 ha) and Devadurga (23,678.92 ha) showing moderate to marginal potential. Maize suitability showed a similar pattern, with Manvi covering 81,745.62 ha (41.5%) of highly suitable land and Devadurga having the most unsuitable area (44,310.65 ha). The assessment identified key biophysical and climatic constraints such as acidic soils (pH 4-5), low soil organic carbon (0-12 g/kg), and limited annual rainfall (440-670 mm), which collectively restrict agricultural potential. Validation with long-term yield records showed strong, statistically significant positive correlations for rice (r = 0.966, p < 0.05) and millet (r = 0.953, p < 0.05), confirming the robustness of the results. These findings provide a spatially explicit foundation for climate-smart interventions and strategic planning in semi-arid Karnataka, offering practical guidance for policymakers, planners, and farmers by identifying soil and climate constraints and informing crop allocation strategies to boost efficiency and productivity.Graphical AbstractThis study utilized an integrated multi-criteria evaluation (MCE) approach, incorporating geospatial analysis, to assess cropland suitability for rice, millet, and maize in the semi-arid sub-districts of Raichur, Karnataka, India. Ten biophysical and climatic variables including rainfall, land surface temperature, slope, soil pH, soil texture, soil organic carbon, and soil moisture along with hydrological parameters such as drainage density and river distance, served as input criteria. A hybrid MCE framework, integrating Fuzzy Analytical Hierarchy Process (Fuzzy-AHP) and TOPSIS, was used to derive factor weights and produce suitability maps in ArcGIS. These maps classified land into four categories: highly suitable, moderately suitable, marginally suitable, and not suitable. Model outputs were validated by correlating land suitability indices (LSI) with long-term yield records (1997-2023), revealing strong, positive, and statistically significant relationships for rice and millet. The resulting suitability maps offer valuable decision support for crop planning and resilience-building in semi-arid agriculture.