Six Lightning Mapping Array (LMA) stations were deployed in April 2010 around Eyjafjallajökull volcano in southern Iceland. Single-station LMA observations were made during the first explosive period (14–18 April), and three-dimensional LMA observations were made during the second explosive period (5–22 May). The single-station observations revealed that continuous RF electrical activity caused by high rates of small vent discharges occurred during the first explosive period, but not the second, indicating that the strength of vent charging varied between the first and second explosive periods. During the second explosive period, very little lightning was detected between 5 and 10 May, while moderate rates of lightning were detected between 11 and 21 May, signaling that another change occurred on 11 May that affected plume electrification. The data do not make clear if it was changing eruptive activity or changing meteorological activity that resulted in the sudden onset of lightning. The plume charge structure during the second explosive period was inferred from the three-dimensional lightning data, showing that the dominant charge structure varied between a positive monopole and a negative-over-positive dipole. The predominance of a low-altitude region of positive charge and the observation that electrical activity was concentrated near the vent indicate that net positive vent charging was dominating the electrification.