Abstract That biodiversity declines with latitude is well known, but whether a metacommunity process is behind this gradient has received limited attention. We tested the hypothesis that dispersal limitation is progressively replaced by mass effects with increasing latitude, along with a series of related hypotheses. We explored these hypotheses by examining metacommunity structure in stream invertebrate metacommunities spanning the length of New Zealand’s two largest islands (∼1,300 km), further disentangling the role of dispersal by deconstructing assemblages into strong and weak dispersers. Given the highly dynamic nature of New Zealand streams, our alternative hypothesis was that these systems are so unpredictable (at different stages of post-flood succession) that metacommunity structure is highly context dependent from region to region. We rejected our primary hypotheses, pinning this lack of fit on the strong unpredictability of New Zealand’s dynamic stream ecosystems and fauna that has evolved to cope with these conditions. While local community structure turned over along this latitudinal gradient, metacommunity structure was highly context dependent and dispersal traits did not elucidate patterns. Moreover, the emergent metacommunity types exhibited no trends, nor did the important environmental variables. These results provide a cautionary tale for examining singular metacommunities. The considerable level of unexplained contingency suggests that any inferences drawn from one-off snapshot sampling may be misleading and further points to the need for more studies on temporal dynamics of metacommunity processes.