Electron beam melting (EBM) is an additive manufacturing process in which fine layers of metallic powder are selectively fused by an electron beam. The process is used to produce parts such as aerospace components and biomedical implants. These applications have driven interest in the development of in-process monitoring, as this could increase confidence in built parts and reduce post-build inspection requirements. In literature, contamination of feedstock has been identified to be problematic for a similar process. Current research efforts are largely limited to infra-red and visual imaging. The aim of this work was to develop monitoring methods which can capture information about topography and contamination during EBM processing, using techniques that are capable of industrialisation. Due to the nature of electron beam interactions with materials, emitted signals carry information about the surface of the irradiated material and its chemical composition. Characteristic x-rays, backscattered electrons (BSE) and secondary electrons (SE) were selected for observation, as these signals are used in scanning electron microscopy for analysis of chemical composition and to capture images. Using an Arcam A1 EBM machine, the conditions in the EBM chamber during building were studied by measuring the temperature in several locations, and the build-up of metallic condensation (metallisation) to understand how they could affect detectors. Process temperatures were found to be prohibitively high for detectors inside the chamber. Metallisation collected from a build with 58 cm3 volume and 20 mm total build height was calculated to allow < 10% transmission of aluminium characteristic x-rays (1.5 keV). Three concepts for low cost x-ray detectors were developed, as high cost detectors used in scanning electron microscopy are difficult to justify in an industrial environment and are likely to require frequent replacement due to the process environment. In ex-situ testing, these detectors were found to be unable to detect x-rays due to their sensitivity to noise. An electron imaging system was developed in the EBM machine. A small area, off-axis electron detector consisting of a metallic mesh in front of a metal plate was used to capture electron images. Use of bias voltage for separating SE and BSE signals for electron imaging was explored; positive biases applied to the mesh reduced current reaching the detector plate, and negative biases increased the current. Despite this, when high bias voltages (200+ V) were used, images showed increased surface sensitivity, indicating SE. Contrast in electron images due to the effect of atomic number on BSE yield was examined. Copper, aluminium, titanium and stainless steel were all found to be identifiable by contrast, and the contrast was clear over a range of bias voltages (±120 V). Negative bias voltages (-60 to -120 V) were tested to see if repelling SE increased responses to materials. It was found to offer some improvement. Detection of small amounts of contaminant materials (aluminium and tungsten) was also explored. 4 µm thick tungsten wire could be seen in electron images captured with 20 µm/pixel. Sintered powder containing contaminant powder particles (aluminium with < 106 µm powder diameter, and tungsten with 45 µm powder diameter) were used to test identification of contaminants, but they could not be identified in electron images. The findings of this research demonstrate the feasibility of in-process quality control by using electron imaging to identify contaminants and abnormalities.