For maximum impact, high yielding improved varieties with significant yield advantages must be targeted to farmers and localities that value this trait most. Explaining farmers’ demand for yield can serve as a means of targeting the development and dissemination of high yielding varieties. This paper analyzes data collected from 395 farmers in northern Ethiopia using a zero-limit Tobit regression. According to the results, poor and marginalized farmers prefer varieties adaptable to poor weather and soils, early maturing, and those which can address diverse concerns than varieties exceptionally good in a single trait (like yield). The richer farmers demand more yield advantage over the existing ones to convince them to use Improved Varieties. For farmers operating in relatively good farming systems (soils, weather, etc.), investment has to be made not only on crop improvement but also on complementary inputs, improved practices, and market development. Farmers who consider improved varieties more marketable and valuable take up high yielding varieties with relatively marginal yield difference. In areas and farmers where there is lower demand for yield, other variety traits (like early maturity, yield stability, and adaptability to local soils/weather) are also important to consider in future crop improvement activities. To ensure that farmers who demand more yield use IVs more productively, the yield advantage, compared to the existing varieties under use, must be high enough and stable.