<p>Farm mechanisation is widely regarded as a key pathway to improve productivity and alleviate rural poverty among smallholder farmers, yet empirical evidence on its multidimensional poverty impacts remains limited. This study analyzes data from 413 households in Ethiopia's Arsi Zone to examine the determinants of mechanisation adoption and its effects on rural multidimensional poverty using the multinomial endogenous switching regression (MESR) model and the Alkire and Foster Rural Multidimensional Poverty Index (R-MPI). Results indicate that access to extension services, credit, and cooperative membership significantly increase the likelihood of adopting both tractors and combine harvesters, while factors such as crop diversification and distance from cooperatives reduce adoption rates. The conditional treatment effects reveal that mechanisation adoption substantially lowers poverty levels: combine harvester use alone reduces R-MPI by 6.75%, tractor adoption alone yields a significant 4.6% reduction, and joint adoption of both machines produces the largest and most significant poverty decrease of 20.51%, equivalent to a 0.09-point decline in R-MPI. These findings highlight the critical role of integrated mechanisation strategies and cooperative support in rural development, suggesting that policies enhancing access to machinery and rural services can effectively reduce multidimensional poverty in Ethiopia and comparable settings.</p>