The early fossil record of terrestrial arthropod herbivory consists of two pulses.The first pulse was concentrated during the latest Silurian to Early Devonian (417 to 403 Ma),and consists of the earliest evidence for consumption of sporangia and stems (and limited fungivore borings). Herbivorization of most of these tissues was rapid, representing 0 to 20 million-year (m.y.) lags from the earliest occurrences of these organs in the fossil record to their initial consumption (Phase 1). For approximately the next 75 m.y., there was a second,more histologically varied origination and expansion of roots, leaves, wood and seeds,whose earliest evidence for herbivorization occurred from the Middle-Late Mississippian boundary to the Middle Pennsylvanian (327 to 309 Ma). The appearance of this second herbivory pulse during the later Paleozoic (Phase 2) is accompanied by major lags of 98 to 54 m.y. between times of appearance of each of the four organ and tissue types and their subsequent herbivory. Both pulses provide a context for three emerging questions. First is an explanation for the contrast between the near instantaneous consumption of plant tissues during Phase 1, versus the exceptionally long lags between the earliest occurrences of plant tissues and their subsequent herbivorization during Phase 2. Second is the identity of arthropod herbivores for both phases. Third is the cause behind the overwhelming targeting of seed-fe plant hosts during Phase 2. Regardless of the answers to these questions, the trace fossil record of plant-arthropod associations provides primary ecological data that remain unaddressed by the body-fossil record alone.