Studies aiming at explaining specialization along latitudinal gradients of plant-herbivore interactions have, to date, yielded inconclusive results. Here we propose the use of steep altitudinal gradients for dissecting factors driving evolution of polyphagy in insect herbivores. First, we test whether colonization of high elevation environment favours increased niche-breadth in two disparate insect groups - the wood-boring beetles and the pollinator group of bees - and show increased polyphagy at higher altitudes in both groups. We then assess classic assumptions transferred from the 'latitude-niche-breadth hypothesis', particularly the increase in environmental variability at high, compared to low, altitude. Finally, we discuss alternative mechanisms shaping the observed pattern of increased polyphagy in altitude, including variation in plant quality and predator pressure at different altitudes. We thus suggest evidence for the 'altitude niche-breadth hypothesis', in which both abiotic and biotic conditions, including increased variability and an increase of the potential feeding niche-breadth, promote evolution for increased insect polyphagy in altitude.