||This paper illustrates obstacles in Participatory Varietal Selection (PVS) through literature review and analyses of data from a questionnaire survey (no = 300), key informant interviews (no = 15), participatory learning trials (no = 6), focus group discussions (no = 6), direct and participant observation. Findings reveal significant pitfalls in PVS that resulted from emergence of spoiler factors such as new crop disease and unpredictable floods. Spoiler factors curtailed sustainability of participatory gains. Besides, learning during critical field stages was thin due to participatory lapses. For instance, an average of only 8% (total = 1488) consistently participated in each key stage of PVS learning trials in western Kenya. The majority (92%) participated on average once, especially in the end-of-season field day when final ranking of crop varieties was done. Consequently, this majority engaged in impulse buying, due to poor grasp of salient productivity factors underlying attractive traits of the new germplasm. Essentially, post-PVS sustainability conditions were elusive. Scaling out of PVS gains was hence compromised and early adopters often fell back to old varieties, to experience, or to sporadic extension advice. PVS therefore has to be anchored within an integrated knowledge exchange system that addresses challenges emanating from interaction of smallholder social, spatial and temporal contexts.