The lecture will begin by a quick review of major overflow locations in the world, along with a description of their characteristics. Then an outline of a typical project to measure flow in the deep ocean will be given. In this case, we measured the flow of Antarctic Bottom Water from the Brazil basin in the western South Atlantic into the Guiana Basin in the western North Atlantic. This water is found below 3900 m depth, and it is confined by a relatively flat 4500-m deep passage between 1 degree S and 4 degrees N with complex sidewall shape. We will show the temperature and salinity signature of this water along with current meter results gathered at two times, once for 22 months in the early 1990’s and later over 44 months in 1998-2002. The conclusion is that this flow is extremely steady and no change in its flux is detected over more than a ten-year time span. There is no global warming signal seen here so far. Next, we will describe a typical laboratory study of this type of flow. Water is pumped in steadily at various locations in a deep rotating upstream basin, with fluid leaving through a level passage. Velocity profiles of currents agree reasonably well with constant potential vorticity theory. To the right of the detached upstream current is a closed gyre that connects the upstream flows (that have different patterns depending on source location) with the unique passage flows. The results suggest that gyres upstream of critically controlling passages in the ocean might serve as adjustment regions between the relatively unconstrained upstream flows and the tightly controlled passage flow. Next we review theory of constant potential vorticity flow, and calculate a velocity profile, a surface height profile, and the total volume flux assuming critically controlled flow at an exit passage. Some features typical of rotating fluids will be shown. Finally the theory is compared to current meter data from all the passages with measurements taken for longer than about a month. We see that comparison is crudely good, and that improving the value of potential vorticity makes it better. As always, there is room for improvement and some possible new directions will be mentioned to conclude the talk.