Inward rectifier K(+) (Kir) channels are expressed in multiple neuronal and glial cells. Recent studies have equated certain properties of exogenously expressed Kir4.1 channels with those of native K(+) currents in brain cells, as well as demonstrating the expression of Kir4.1 subunits in these tissues. There are nagging problems however with assigning native currents to Kir4.1 channels. One major concern is that in many native tissues, the putatively correlated currents show much weaker rectification than typically reported for cloned Kir4.1 channels. We have now examined the polyamine-dependence of Kir4.1 channels expressed at high density in Cosm6 cells, using inside-out membrane patches. The experiments reveal a complex and variable rectification that can help explain the variability reported for candidate Kir4.1 currents in native cells. Most importantly, rectification seems to be incomplete, even at high polyamine concentrations. In excised membrane patches, with high levels of expression, and high concentrations of spermine, there is approximately 15% residual conductance that is insensitive to spermine. From a biophysical perspective, this is a striking finding, and indicates either that a bound spermine fails to completely block permeation or that significant spermine permeation (i.e. 'punchthrough') is occurring. To examine this further, we have examined block by philanthotoxin (PhTx, essentially spermine with a bulky tail). PhTx block, while less potent, is more complete than spermine block. This leads us to propose that spermine 'punchthrough' may be significant in Kir4 channels, and that this may be a major contributor to the weak rectification observed under physiological conditions.