This paper intends to utilise magnetic resonance (MR) data to predict the rock-type (RRT), by correlating its profile to the Mercury Injection Capillary Pressure (MICP), using a novel semi-automated method. Data presented come from Oilfield A, a reservoir highly affected by diagenesis causing reservoir properties to vary laterally. As a first step the feasibility of the method will be tested on the pilot well dataset, where its suitability will be assessed before applying the technique to the development well MR dataset. The fieldwide application in both pilot and development wells will improve the confidence of rock-type distribution. Two methods, normalised cumulative and cumulative are attempted, correlating the inferred NMR pore volume to the inferred MICP pore throat distribution, grouped according to the rock type scheme in two reservoirs. Both NMR and MICP data are arranged in the same format. On the normalised cumulative method, feasibility is conducted on MICP data to see at which normalised cumulative frequence the rock-type shows distinct pore throat clusters, while on the cumulative method the rock-type clustering is assessed on the cumulative frequency range at the total pore throat size. A key assumption is that the NMR relaxation profile representing the pore size, is similar to the MICP pore throat profile to classify the rock-type. QC is performed on the resulting curves to ensure they honour the reservoir property trends from core. One method will be chosen on this basis, where a set of criteria will be applied to the resulting curve to generate the first pass rock-type prediction. This paper contains the step 1 of rock-type estimation in the pilot wells and the suitability of application to the development wells. The development well application itself is currently still ongoing.