Bamboo is the fastest growing plant with a relatively short rotation period and high biomass yield. However, there are little practical experience or documented traditional knowledge on how local communities manage, process and sustainably utilize bamboo resource. Therefore, the present study was aimed at: (1) assessing and documenting ethnobotanical knowledge in Chencha, Hula and Masha Districts; (2) identifying major predictor variables that influence the traditional knowledge; and (3) develop a Generalized Linear Model (GLM) to predict or explain the traditional knowledge on highland bamboo. A total of 288 informants comprising 169 men and 119 women were involved between the ages of 18 and 99. Data were collected using structured interviews, discussions and field observation. Our findings showed that 98% of the informants planted highland bamboo. The resource covered 0.93 ha (18.2%) of the total land area per household. Its harvesting contributed 17.4% cash income compared to honey (0.6%) and Eucalyptus (0.5%). Unlike Chencha and Hula Districts, bamboo shoot processing and consumption as food supplements in Masha District was a common practice. About 69% of the informants’ propagated highland bamboo using mother stock and 92.7% used cattle dung and ash. Prediction of GLM assured informants’ age and gender, time of seedling plantation and season of culms harvesting significantly (p < 0.001) influenced traditional knowledge. Lack of awareness and training and limited market access were some of the bottlenecks. Documenting traditional knowledge can serve as a basis for bamboo management, processing as well as sustainable utilization.