Abstract We constrain the distribution of spatially offset Lyman-alpha emission (Ly α) relative to rest-frame ultraviolet emission in ∼300 high redshift (3 < z < 5.5) Lyman-break galaxies (LBGs) exhibiting Ly α emission from VANDELS, a VLT/VIMOS slit-spectroscopic survey of the CANDELS Ultra Deep Survey and Chandra Deep Field South fields (≃0.2 deg2 total). Because slit spectroscopy only provides one spatial dimension, we use Bayesian inference to recover the underlying two-dimensional Ly α spatial offset distribution. We model the distribution using a two-dimensional circular Gaussian, defined by a single parameter σr,Ly α, the standard deviation expressed in polar coordinates. Over the entire redshift range of our sample (3 < z < 5.5), we find $σ _{r,\mathrm{Ly}\,α }=1.70^{+0.09}_{-0.08}$ kpc ($68\hbox{ per cent}$ conf.), corresponding to ∼0${^{\prime\prime}_{.}}$25 at 〈z〉 = 4.5. We also find that σr,Ly α decreases significantly with redshift. Because Ly α spatial offsets can cause slit losses, the decrease in σr,Ly α with redshift can partially explain the increase in the fraction of Ly α emitters observed in the literature over this same interval, although uncertainties are still too large to reach a strong conclusion. If σr,Ly α continues to decrease into the reionization epoch, then the decrease in Ly α transmission from galaxies observed during this epoch might require an even higher neutral hydrogen fraction than what is currently inferred. Conversely, if spatial offsets increase with the increasing opacity of the intergalactic medium, slit losses may explain some of the drop in Ly α transmission observed at z > 6. Spatially resolved observations of Ly α and UV continuum at 6 < z < 8 are needed to settle the issue.