Levitation techniques have been applied to a staggering range of materials, from liquid helium to aqueous solutions of proteins, to metals, ceramics, glasses and semiconductors. These experiments have encompassed temperatures from cryogenic to greater than 2500 °C, and samples from micrograms to tens of kilograms. It should come as no surprise that a wide variety of levitation principles have been employed for processing these samples, including electromagnetic (EML), electrostatic (ESL), aerodynamic and gas film, acoustic and dia- or paramagnetic levitation, as well as combinations of these and others. All of these containerless techniques share one key feature: internal flow in liquid samples. The flow may be driven directly by the positioning force, as in EML and aerodynamic levitation, or by temperature gradients through Marangoni convection and/or natural buoyancy, as in ESL. It is possible to reduce the positioning-driven and buoyancy flows by performing the experiments in microgravity; however, often even the reduced levels are important. For some experiments, such as viscosity measurements, only whether the flow is laminar or turbulent must be established. For other experiments, however, quantitative assessments of velocity, shear stress or shear strain rate are required. In most cases, it is difficult or impossible to measure the internal flow in levitated droplets directly. The samples are usually small, and often opaque, reactive, high-temperature, metastable, or all of the above. Furthermore, recirculating flow limits the utility of tracking surface particles, since they tend to collect in stagnation points rather than following the flow. Most research groups have chosen mathematical modelling to assess the internal flow in levitated droplets. Several different classes of experiments are examined in terms of the effect of fluid flow and the impact of flow modelling. This paper focuses on EML and ESL, although the techniques and many of the results are applicable to other levitation methods.