It is widely recognized that the growth of new (juvenile) masses of continental crust (from all sources) is estimated at approximately 1.7 km (super 3) /yr. Geological and geophysical observations at convergent ocean margins imply that processes of sediment subduction (i.e., underthrusting of ocean floor sediment to mantle depths) and subduction erosion (i.e., removal of upper plate rock to mantle depths) recycles terrestrial matter (i.e., continents and island arcs) to the mantle at a similar rate. Over long periods of geologic time (10s to 100s my), these twained non-accretionary processes return an estimated minimum of 0.7 and 1.0 km (super 3) /yr of terrestrial matter, respectively, to the mantle reservoir. During the past half-decade the combination of an improving inventory of observations at Pacific subduction zones, in particular where supported by the Ocean Drilling Program (ODP--e.g., at Tonga, Chile, Costa Rica), and theoretical and laboratory modeling results argue that the long-term rates of sediment subduction and subduction erosion are, in fact, higher than the minimum estimates. For example, drilling has determined that 100 percent sediment subduction takes place off Coast Rica, and that subduction erosion landward of this sector of the Middle America Trench maybe be 2-3 time the estimated minimum global rate, which is 30 km (super 3) /my/km of trench axis. Because many problems vex quantitative calculations, in particular the lack of age and rock control that ODP drilling can uniquely provide, it can only be inferred that the minimum rates of crustal recycling should be increased to at least 2.0-2.5 km (super 3) /yr. Similarly, the rock inventory of the central Aleutian arc, which has remained a coherent igneous and sedimentary body for roughly 50 my, and to which is added-back arc material removed by processes of sediment subduction and erosion, imply arc-production rates near 100 km (super 3) /my/km of arc. This is much higher than common estimates (long term) of arc magmatism, which is the major contributor to continental growth. So, whereas the global rate of crustal recycling is doubtlessly higher than current models state, the same seems to be true for growth of arc-magmatic piles. A key term involved in calculating both rates is crustal or upper plate losses arising from subduction erosion, which probably accounts for 3/4 of all refluxed materials. Based on observations at modern convergent margins, it can be inferred that a general balance may exist between the formation and recycling of terrestrial crust.