Different scholars have modelled the adoption of sustainable agricultural practices (SAPs) with a goal of improving farmer's adaptive capacity to climate change. Nonetheless, through the conventional way of defining adoption decisions as one-time survey decisions, many scholars have failed to understand inconsistencies in adoption decisions and dis-adoption of such practices. Through a survey of 2100 maize farming households, the current study employed multivariate probit models to understand and compare one-time survey season adoption decisions and sustained (consistent) adoption decisions. The study notes that dis-adoption rates of SAPs range from 20 to 27 percent. As such, the determinants of dis-adoption were estimated to build a case for going beyond one-time adoption survey decisions. Furthermore, the study employed a Cox Proportional hazard model to understand the relative risk to adoption of Sustainable Agricultural Practices over time. The findings reveal the need for a modelling paradigm shift in understanding adoption decisions for sustainable benefits. Past research has focused on understanding whether farmers adopted SAPs in a previous survey season period. Nonetheless, efforts to ensure sustainable productivity and food security remains lacking amidst climate change. The current study reveals evidence of a need for a modelling shift towards sustained adoption of SAPs for sustainable development of agriculture. Lastly, the findings reveal the need for intensifying knowledge and information dissemination on SAPs through field demonstrations, extension visits, trainings and radio programs in order to reduce dis-adoption and ensure sustained adoption.