The purpose of this study was to describe the locomotor activity of men's hockey based on the phase of play and to identify if differences in locomotor activity exist between phases. Twenty-four players (age 26 ± 4) wore a 10 Hz Global Positioning System device to track locomotor activity during 10 international matches. The locomotor activity of players was allocated to five different phases: (1) established attack, (2) opposition counterattack, (3) established defence, (4) attacking counterattack and (5) offensive pressing, utilising a video analysis-based system. Linear mixed models and post-hoc pairwise comparisons, using estimated marginal means, were utilised to compare the locomotor activity in each phase of play within each position. Significant main effects were found for the phase of play on several locomotor activity metrics ( p ≤ 0.05). With respect to max speed and relative total distance across positions, counterattacks present a unique challenge compared to other phases. Established attack phases evoke lower max speed values than offensive pressing phases for all playing positions as well as compared to opposition counterattacks for defenders, midfielders, and forwards. Positional differences existed during attacking counterattack phases, with defenders producing lower values than both midfielders and forwards for high-speed running, high-speed running efforts, max speed and relative total distance. Depending on the metric, this approach captured 22%–70% of a player's locomotor activity providing valuable insight into the current dynamics of international hockey. This investigation demonstrated positional differences based on the phase of play, which may explain previous research findings regarding positional differences.