This study examines the presence of herding behaviour in the cryptocurrency market. The latter is the outcome of mass collaboration and imitation. Results from the static model suggest no significant herding. However, the presence of structural breaks and nonlinearities in the data series suggests applying a static model is not appropriate. Accordingly, we conduct a rolling-window analysis, and those results point to significant herding behaviour, which varies over time. Using a logistic regression, we find that herding tends to occur as uncertainty increases. Our findings induce useful insights related to portfolio and risk management, trading strategies, and market efficiency.