Accurately quantifying marine carbon export and sequestration is a prerequisite not only for the development of reliable predictions of the ocean’s capacity to store anthropogenic carbon, but also a requirement for the future allocation of carbon credits.  Although gravitational settling is thought to contribute most to carbon sequestration, particle sinking velocities remain poorly constrained and represented in models. Estimates of export flux based on particle size–sinking velocity relationships have been shown to be unreliable, and associated carbon supply versus demand budgets for the upper mesopelagic do not provide sufficient carbon, revealing conceptual misconceptions. Marine snow catchers, which separate marine particles into sinking versus non-sinking ones, revealed that the vast majority of particulate organic carbon of the mesopelagic must be considered effectively non-sinking.  This challenges our traditional concepts that highlight, below the ocean surface, the sinking carbon. Using the KISS principle (Keep It Simple, Stupid), we conducted a meta-analysis of marine snow catcher data from the upper mesopelagic to explore if biochemical composition, in concert with estimates of marine snow concentrations, provides a better understanding of carbon flux. We compared the relative contributions of particulate organic carbon, ballasting material (biogenic silica, lithogenic silica, calcium carbonate) and buoyant exopolymers (e.g. transparent exopolymer particles and Coomassie stainable particles) between sinking and suspended particles in the upper mesopelagic (200-500 m). Suspended particle composition consistently differed from that of sinking particles.  Further, the differences depended on ecosystem state ranging from oligotrophic conditions to a diatom or Phaeocystis bloom. We find that exopolymers, which are buoyant, may efficiently counteract the effect of ballasting material. A relative lack of exopolymers may explain the difference between sinking and suspended particles under some non-bloom conditions, whereas large, ballasted marine snow particles likely caused rapid sinking at the end of a diatom bloom. Dynamics were complex during a Phaeocystis bloom, where large Phaeocystis snow particles initiated sinking, but became buoyant due to the large contribution of exopolymers.