We present a measurement of the volumetric rate of superluminous supernovae (SLSNe) at z ∼ 1.0, measured using archival data from the first four years of the Canada–France–Hawaii Telescope Supernova Legacy Survey (SNLS). We develop a method for the photometric classification of SLSNe to construct our sample. Our sample includes two previously spectroscopically identified objects, and a further new candidate selected using our classification technique. We use the point-source recovery efficiencies from Perrett et al. and a Monte Carlo approach to calculate the rate based on our SLSN sample. We find that the three identified SLSNe from SNLS give a rate of $$91^{+76}_{-36}$$ SNe yr^−1 Gpc^−3 at a volume-weighted redshift of z = 1.13. This is equivalent to 2.2$$^{+1.8}_{-0.9}\times 10^{-4}$$ of the volumetric core-collapse supernova rate at the same redshift. When combined with other rate measurements from the literature, we show that the rate of SLSNe increases with redshift in a manner consistent with that of the cosmic star formation history. We also estimate the rate of ultra-long gamma-ray bursts based on the events discovered by the Swift satellite, and show that it is comparable to the rate of SLSNe, providing further evidence of a possible connection between these two classes of events. We also examine the host galaxies of the SLSNe discovered in SNLS, and find them to be consistent with the stellar-mass distribution of other published samples of SLSNe.