The longitudinal profile of the 320 EeV event observed by the Fly's Eye experiment is analysed. A method of testing the hypothesis of a specific primary particle type is described. Results for different particle types are summarized. For hadronic primaries between proton and iron nuclei, the discrepancy between observed and simulated profiles is in the range of 0.6-1.0 sigma for two different hadronic interaction models investigated. For primary photons, the discrepancy is 1.5sigma assuming a standard extrapolation of the photonuclear cross-section with energy. Larger values of the cross-section at highest energies make primary photon showers more similar to hadron-initiated events. The influence of varying the extrapolation of the photonuclear cross-section is studied.