The work carried out in this paper assessed how processing conditions and feedstock affect the quality of the coke produced during microwave coke making. The aim was to gather information that would support the development of an optimised microwave coke making oven. Experiments were carried out in a non-optimised 2450 MHz cylindrical cavity. The effect of treatment time (15-120 min), power input (750 W-4.5 kW) and overall power input (1700-27,200 kWh/t) on a range of coals (semi-bituminous-anthracite) was investigated. Intrinsic reactivity, random reflectance, strength index and dielectric properties of the produced cokes were compared with those of two commercial cokes to assess the degree of coking produced in the microwave system. Overall energy input and coal rank were found to be the major factors determining the degree of coking following microwave treatment. The dependency on coal rank was attributed to the larger amount of volatiles that had to be removed from the lower ranked coals, and the increasing dielectric loss of the organic component of the coal with rank due to increased structural ordering. Longer treatment times at lower powers or shorter treatment times at higher powers are expected to produce the same degree of coking. It was concluded that microwave coke making represents a potential step-change in the coking industry by reducing treatment times by an order of magnitude, introducing flexibility and potentially decreasing the sensitivity to quality requirement in the feedstock. The main challenges to development are the energy requirements (which will need to be significantly reduced in an optimised process) and penetration depth (which will require an innovative reactor design to maximise the advantage of using microwaves). Understanding and quantifying the rapidly changing dielectric properties of the coal and coke materials is vital in addressing both of these challenges. (C) 2014 Elsevier B.V. All rights reserved.