The study examines the technical efficiency and adoption dynamics of integrated smart farming system in Erode district of Tamil Nadu. Smart farming is represented as a bundled, firm-supported system integrating drip irrigation, fertigation and real-time digital advisory. The primary data were gathered from 120 farmers (60 adopters and 60 non-adopters) in 2025–26 (agricultural year) with stratified random sampling design. Data envelopment analysis (DEA) in variable returns to scale (VRS) and bias corrected using Simar and Wilson bootstrapped DEA were employed to estimate technical efficiency. Bootstrap truncated regression was used to analyse determinants of efficiency and a binary Logit model was used to analyse adoption behaviour. The outcomes revealed a mean technical efficiency of 0.62, representing a 38 % unrealised production potential. Fertiliser accounted for 93.30 % of the input slack, indicating mismanagement as a potentially important source of inefficiency. Education and technology adoption intensity are found to be significant drivers of efficiency. The extension linkage, perceived usefulness and technological awareness are key factors in the decision to adopt the smart farming technologies. The results underscore the need for the adoption of integrated technology packages, capacity development of farmers and extension support to improve the efficiency and productivity of farms.