Drill cores through modern coral reefs commonly show a time lag in reef initiation followed by a phase of rapid accretion to sea level from submerged foundations – the so-called ‘catch-up response’. But because of the difficulty of drilling in these environments, core distribution is usually restricted to accessible areas that may not fully represent reef history, especially if the reef initiated in patches or developed with a prograde or retrograde geometry. As a consequence, core data have the potential to give a misleading impression of reef development, particularly with respect to the timing of initiation and response to sea-level rise. Here, we use computer models to simulate keep-up reef development and, from them, quantify variations in the timing of reef initiation and accretion rate using mock cores taken through the completed simulations. The results demonstrate that cores consistently underestimate the timing of reef initiation and overestimate the reef accretion rate so that, statistically, a core through a keep-up reef will most likely produce a catch-up pattern – an initiation lag followed by a phase of rapid accretion to sea level. This implies that catch-up signatures may be an artefact of coring and that keep-up reefs are significantly more common than previous core studies claim.