This study examines the temporal (monthly) and spatial climatology (2004–2010) of the first return stroke of the cloud‐to‐ground (CG) lightning flash peak current (Ip) across various land/water boundaries over the contiguous United States. Four regions are examined: the Gulf of Mexico (region 1), the Florida peninsula (region 2), Lake Michigan (region 3), and part of the U.S. Mid‐Atlantic (region 4). The crosss across the coastlines of regions 1, 2, and 4 show a gradual oceanward increase in the mean negative polarity CG peak current values (−Ip). This transition along the respective land/ocean boundaries is not sharp but gradual. In direct contrast with ocean, there is no consistent behavior in −Ip values as we move from land out across the fresh water of Lake Michigan (region 3). Meanwhile, the positive CG flash peak current (+Ip) values do not exhibit a consistent variation across any coastal boundary. For region 1, the −Ip values increase as we move toward the coast (southwards) especially during the wet season (June–October). This finding is in direct contrast with studies that documented winter as the season of maximum −Ip values. The zonal and seasonal variations of −Ip values across region 4 are not quite as pronounced, but the oceanic −Ip values are still larger than over the adjoining landmass. We explore in turn which up to date hypotheses pertinent to the oceanic −Ip enhancement are supported or refuted by our findings. It is concluded that the oceanic −Ip enhancement is not an artifact related to CG detection or Ip retrieval methods, nor is it likely related to the cloud top heights or CG activity. The study cannot refute the role of electrical conductivity and its contribution to CG leader attachment processes. However, given the observed “blurred transition” of the Ip values across the coastlines this paper suggests that likely the main physical mechanism is acting on the thundercloud potential. The recently suggested role of sodium chloride (NaCl) but also the role of ice crystal size (implicated herein), as possible modulators of the thundercloud potential, exhibit distinct pros and cons. Their candidacy is supported by their strong physical links to the electrostatic charging and thundercloud electric potential buildup but also by the exhibited blurred −Ip transition across the coastlines. In contrast, the suggested mechanisms cannot individually explain the observed −Ip enhancement in terms of season, NaCl concentrations, and absence of similar behavior in the respective +Ip values.