Forest structure and composition play an essential role in determining the carbon storage capacity of tropical forests. Andean forests, with great potential for carbon accumulation, include large expanses of high-density woody bamboo communities. Woody bamboos can potentially alter forest structure, composition and dynamics and thus can affect carbon storage capacity; however, they are commonly excluded from forest monitoring and modelling. With the aim of documenting patterns of bamboo abundance and disentangling its association with forest structure, we carried out a bamboo census in seven 1-ha long-term forest monitoring plots situated across a large elevation gradient (1000–3600 m a.s.l.) in the Peruvian Andes. We determined that bamboo is a dominant plant group in the study area. In every plot, bamboos were the most common genera in terms of number of stems, and in two of the plots bamboo species were among those with the greatest basal area. We used a combination of Generalized linear mixed models (GLMM) and structural equation modelling (SEM) to hypothesize a causal framework and determine the direction and size of the effects of bamboo abundance (basal area) on number of individual trees, total tree basal area, mean tree basal area, mean tree growth rate and tree mortality rate. We found an overall negative association between bamboo abundance and total tree basal area driven mainly by reduced tree density (directly and indirectly mediated by an increase in tree mortality). However, the decrease in tree density and the increase in tree mortality are also associated with a small increase in tree diameter (mean tree basal area). Overall, the negative association between bamboo abundance and tree basal area suggests a lower biomass accumulation and thus a lower carbon storage capacity of trees in Andean forests where bamboo is dominant. Our results, which show the importance of bamboo in determining forest function, highlight the need for including bamboo in monitoring efforts and modeling studies.