A major difficulty in managing wildlife trade is the reliance on trade data (rather than capture data) to monitor exploitation of wild populations. Collected organisms that die or are rejected before a point of sale often go unreported. For the global marine aquarium trade, identifying the loss of collected fish from rejection, prior to export, is a first step in assessing true collection levels. This study takes a detailed look at fish rejections by buyers before export using the Papua New Guinea marine aquarium fishery as a case study. Utilizing collection invoices detailing the species and quantity of fish (Actinopteri and Elasmobranchii) accepted or rejected by the exporting company it was determined that, over a six month period, 24.2% of the total fish catch reported (n = 13,886) was rejected. Of the ten most collected fish families, rejection frequency was highest for the Apogonidae (54.2%), Chaetodontidae (26.3%), and Acanthuridae (18.2%) and lowest for Labridae (6.6%) and Hemiscylliidae (0.7%). The most frequently cited reasons for rejection were fin damage (45.6% of cases), undersized fish (21.8%), and fish deemed too thin (11.1%). Despite fishers receiving feedback on invoices explaining rejections, there was no improvement in rejection frequencies over time (r = -0.33, P = 0.15) with weekly rejection frequencies being highly inconsistent (range: 2.8% to 79.4%; s = 16.3%). These findings suggest that export/import statistics can greatly underestimate collection for the marine aquarium trade as additional factors such as fisher discards, escapees, post-collection mortalities, and unregulated domestic trade would further contribute to this disparity.