Networks play a key role in the functioning of socioecological fishery systems. Most network studies among fish harvesters examining fishing success utilize interviews and questionnaires. Though insightful, such studies are resource and time-intensive and thus unlikely to be replicated frequently through time. Alternatively, commercial landings records and vessel monitoring systems (VMS) provide continuous sources of information that can be used to examine variation in vessel networks through time. We used VMS data to define association networks among vessels. Relationships were found between common network metrics and annual performance based on landings data. Associations between vessels were more closely examined as a function of annual activity, performance, favoured species, and landing port using temporal exponential random graph models. We examined network dynamics across 4 consecutive years. Changes in vessel associations were clearly related to performance, landing port, and species targeted. Network structure could affect the relationship between catch and nominal effort, influencing stock assessments and responses to management actions. Our methodology provides a means to follow network change, identifying situations where more detailed study is warranted.