Two distinctly different types of fish hooks are reported in literature, of which, the most common shape shows a decrease in efficiency to a minimum followed by a monotonic increase with size. The other shape exhibits an increase in efficiency with size till a critical point is reached; the efficiency then decreases till a minimum is reached, followed by a monotonic increase with size. Early theories to explain fish hook phenomenon through variable bypass are shown to be simple mathematical transformations. The 'mechanistic model' attributes occurrence of fish hook to a sharp fall in settling velocities in a centrifugal field with change in flow from Stokesian to transient regime. This explanation is shown to be dubious and not in conformity with known principles of physics. Also, the boundary layer model and the entrainment model require considerable refinement as mechanics of fluid flow around irregular particles is not well developed. It is shown that CFD simulations show efficiency curves with or without fish hook effect depending upon the assumptions for simulation. The method of sizing analyses in detection of fish hook effect is critically discussed. Most of the occurrences of fish hook are reported when sizing analyses are carried out by laser diffractometry using Fraunhofer approximation for data interpretation. When alternate sizing techniques, such as Andreasen pipette, disc centrifuge, Coulter counter, Dynamic Light Scattering etc. are used or when Mie theory is applied in Laser techniques, fish hook is not reported. Fish hook effect appears to be repeatable where it was studied. However, the conditions of reproducibility are to be specified by the proponents, if the phenomenon is to find general acceptance. It is likely that fish hook would continue to be regarded as a placebo. Its exclusion in simulation models does not appear to affect cyclone performance prediction. (C) 2014 Elsevier B.V. All rights reserved.