Dissolved CH4 was measured in coastal waters of the southern North Sea, in two adjacent U.K. estuaries with well-defined turbidity maxima (Humber and Tyne) and in their associated river catchments, during a series of campaigns covering the period 1993--1999. In general, samples from all three environments were significantly to highly CH4 enriched relative to atmospheric air. Observed river water concentrations, ~33--152 nmol L-1 (940--4305% saturation) for the Humber river catchment and ~3--62 nmol L-1 (86--1754% saturation) in the river Tyne, were within but toward the low end of the range of CH4 concentrations in river waters world wide. In sea waters from the outer Wash estuary (U.K. coast) and adjacent to the Dutch coast, CH4 was highly but nonlinearly correlated with salinity, consistent with strong CH4 removal from river and/or estuarine CH4 sources influencing these locations. In transects along the Humber and Tyne estuaries, CH4 was highly negatively nonconservative, confirming the estuarine removal hypothesis. For both estuaries, highest CH4 concentrations, ~190--670 nmol L-1 (6000--21,000% saturation) in the Humber and ~650 nmol L-1 (21,800% saturation) in the Tyne, were observed at very low salinity in the vicinity of the turbidity maximum. Importantly, these concentrations greatly exceeded measured river water values, implying for both situations the existence of a large in situ CH4 supply associated with high turbidity. Time series measurements at two locations in the upper Tyne subsequently confirmed the strong correspondence of dissolved CH4 and turbidity in the vicinity of the turbidity maximum. CH4 removal estimated for the Humber, Tyne, Wash, and Rhine-Scheldt estuaries was ~90% of the low-salinity CH4 input. On the basis of this and river discharge data, ~7.1¿108mol CH4 may be removed annually in estuaries bordering the southern North Sea. Of this, ~6.6¿108mol may be lost by air-sea exchange. This represents an additional atmospheric CH4 flux from the North Sea unaccounted for in previous work, which may have, consequently, underestimated this source by ~50%. Upward scaling of this estimate based on the mean of reported river water CH4 concentrations implies a previously unaccounted for ~6.3--24¿109mol (i.e., ~0.1--0.4¿1012g) CH4yr-1 which may be lost globally to gas exchange in estuaries, increasing previous such estimates by ~8--50%. However, as it is based on data that exclude the possibility of elevated CH4 levels at estuarine turbidity maxima, even this revision is likely to be conservative. Detailed studies of CH4 distributions in major world estuaries would now be required in order to successfully reevaluate the CH4 budget of the coastal marine atmosphere. ¿ 2000 American Geophysical Union