Attraction and passage efficiency were reviewed and compared from 19 monitoring studies that produced data for evaluations of pool-and-weir, Denil, vertical-slot and nature-like fishways. Data from 26 species of anadromous and potamodromous fishes from six countries were separated by year and taxonomic family into a matrix with 101 records. Attraction performance was highly variable for the following fishway structures: pool-and-weir (attraction range = 29–100%, mean = 77%, median = 81%), vertical-slot (attraction range = 0–100%, mean = 63%, median = 80%), Denil (attraction range = 21–100%, mean = 61%, median = 57%) and nature-like (attraction range = 0–100%, mean = 48%, median = 50%). Mean passage efficiency was inversely related to mean attraction efficiency by fishway structure type, with the highest passage for nature-like fishways (range = 0–100%, mean = 70%, median = 86%), followed by Denil (range = 0–97%, mean = 51%, median = 38%), vertical-slot (range = 0–100%, mean = 45%, median = 43%) and pool-and-weir (range = 0–100%, mean = 40%, median = 34%). Principal components analysis and logistic regression modelling indicated that variation in fish attraction was driven by biological characteristics of the fish that were studied, whereas variation in fish passage was related to fishway type, slope and elevation change. This meta-analysis revealed that the species of fish monitored and structural design of the fishways have strong implications for both attraction and passage performance, and in most cases, existing data are not sufficient to support design recommendations. Many more fishway evaluations are needed over a range of species, fishway types and configurations to characterize, to optimize and to design new fishways. Furthermore, these studies must be performed in a consistent manner to identify the relative contributions of fish attraction and passage to overall fishway performance at each site.