Molecular shape and electrostatic distribution play a crucial role in enzyme and receptor recognition and contribute extensively to binding affinity. Molecular similarity and bioisosterism are much-discussed topics in medicinal chemistry. Many molecular representations and similarity metrics are available to help drug discovery, and activities such as compound hit explosion and library design can be undertaken using them. The quality of the resulting compound series is highly dependent upon the molecular representation and similarity metric used. We have used a range of software to investigate whether molecules can be represented and compared effectively using measures of three-dimensional shape and electrostatic distribution ("electroforms"). We find that these descriptors allow for the assessment of molecular similarities using standard molecular visualization tools and offer a method for comparing molecules that may be considered superior to other methods.