Since adoption is a dynamic process that involves learning about new technologies, static adoption models fail to adequately explore the effects of changes in farmers’ perception and attitudes over time. This study analyzed the influences of farmers’ learning and risk on the likelihood and intensity of adoption of improved tef and wheat technologies in Northern and Western Shewa zones of Ethiopia. The study employed Xtprobit and Xttobit and random effect models and panel data of the same farmers from 1997 to 2001. Separate samples were selected for wheat and tef and the study covers the same farmers from 1997-2001. Panel data are better suited to study dynamic changes and the random effect models control for unobserved variability and potential endogeniety. Comparison of the main features of tef and wheat farmers revealed that wheat farmers are slightly younger, more educated, have slightly higher family size and significantly higher family labour than tef farmers. While average farm size is similar for tef and wheat farmers, farmers cultivated 60% and 30% of their land to tef and wheat, respectively. However, tef farmers allocated only 20% of their tef area to improved varieties due to shortage of desirable varieties whereas wheat farmers allocated 90% of their land to improved varieties from 1997 to 2001. Only three improved varieties were demonstrated and limited quantities of improved seeds were distributed to tef farmers whereas six improved wheat varieties were demonstrated and relatively sufficient quantities of improved seeds were distributed to wheat farmers during the study. Besides, similar levels of fertilizers and herbicide were used on tef and wheat. Wheat and tef were mainly grown for own consumption as less than half of the produce (48% of all wheat and 46% of all tef) was sold in the market. The study provided evidence of the importance of learning in the adoption decision and area allocation to improved varieties. As farmer’s gained more experience from growing the new varieties in previous years, they continued adoption and increased areas under these varieties. The study also revealed that adopters of wheat and tef technologies have increased their production by 20% and 39%, respectively, than non-adopters. Results of the analyses indicate that awareness, availability and profitability of the new improved tef and wheat varieties enhanced farmer’s learning and farmer’s experience had positive influence on the likelihood and intensity of improved seed adoption. Improved tef and wheat varieties were found more risky than the local varieties. The study further revealed that younger age of farmer, farmers’ learning from previous experience, availability of family labour and credit are key determinants of the likelihood and intensity of adoption of improved seed. Policies and strategies that contribute to timely availability of improved inputs and provision of credit enhance farmers learning from their own experience on adoption. Policies and strategies that focus on farmers’ education and provision of insurance for crop failure to reduce risk would help the new extension program (NEP) achieve its objectives which give emphasis to raising smallholders’ production and productivity.