The rollout of advanced metering infrastructure, advanced distribution automation schemes, and integration of generation into distribution networks, along with a raising of awareness of power quality (PQ), means that there is an increase in the availability of power system monitoring data. In particular, the data for harmonics, whether it is voltage or current harmonics, is now available from a large number of sites and from a diverse range of PQ instruments. The traditional analysis and reporting of power quality examines harmonic orders to the 50th. This means that the harmonic data available for analysis are significantly larger than, for example, steady-state voltage variations where only a few parameters are examined (e.g., the voltage on each phase). Higher frequency components, sometimes called high-frequency harmonics, in the 10-250 kHz range arising primarily due to power-electronic interfaced generation are also becoming significant. Given the vast amount of harmonic data that will be captured through grid instrumentation, a significant challenge lies in developing methods of analysis and reporting that reduces the data to a form that is easily understood and clearly identifies issues but does not omit important details. This paper introduces a number of novel methods of analysis and reporting which can be used to reduce vast amounts of harmonic data for individual harmonic orders down to a small number of indices or graphical representations which can be used to describe harmonic behavior at an individual site as well as at many sites across an electricity network. The methods presented can be used to rank site performance in order for mitigation strategies. The application of each method described is investigated using real-world data.