Particle settling velocity is a critical operating parameter in many industrial processes. Several correlations with reasonable accuracy exist for predicting settling velocity for non-spherical particles: however, they all share a common shortfall: they require detailed 3D knowledge of the particle shape and size that is not available in many practical cases, particularly in an on-line industrial context. These correlations are typically therefore unsuitable for use in situations where limited particle geometric information is available or where particle characteristics can change during process operation. This paper presents a method of predicting settling velocity and drag coeffecient with only 2D geometric information that can be obtained from a single side view of a particle, which is feasible to obtain using on-line imaging techniques. The correlation is formed using a large set of data obtained from the literature. We show that for a set of irregular volcanic particles not used to form the correlation, 74% of the predictions are within +/- 25% error. For comparison, when a standard correlation that uses 3D geometric information is applied to the same set of particles, 81% of predictions are within that error range. The results show that it is possible to obtain predictions using only side-view geometric data with an accuracy close to that achieved by other correlations that require, at minimum, particle surface area and volume. The technique works well when particles have aspect ratios less than five, a range that includes most particle shapes encountered in nature and in industrial processes. The new correlation presented here enables the prediction of settling velocity with reasonable accuracy in an on-line context for many industrial processes, which has not previously been possible. (C) 2018 Published by Elsevier B.V.