Eighty-one consecutive patients representing a wide clinical spectrum were studied to assess the degree of limitation of the atrial extrastimulus technique for measurement of atrioventricular node refractoriness; and, to determine the relationship between conductivity and functional refractoriness of the atrioventricular node (AVN). Four patients developed sustained atrial fibrillation and a total of 277 basic cycle lengths (BCLs) were evaluated in the remaining 77. Four different patterns of AVN conduction were defined on the basis of plots of H1H2 versus A2A2 and discontinuity in the plot was observed for some of each of the three major patterns. No functional refractory period (FRP) of AVN could be determined in 10/77 patients (13%) and in 127/277 BCLs (46%). A mathematical formula relating measurements of conductivity (A2H2 interval) to FRP which was derived from the canine heart was applied to the data for those BCLs where FRP was known. Optimal agreement between measured and calculated FRP (y = x + 3, r = 0.97, n = 38) was obtained when subsets of data were selected according to an algorithm derived from repetetive computer analysis of the parent data base. The calculation is independent of patient variables (medication, impaired AVN conduction, pattern of AVN conduction). These results demonstrate that the atrial extrastimulus technique often will not permit measurement of AVN refractoriness in the human heart but conductivity and FRP are related in a fashion similar to that determined for the canine heart. This relationship permits calculation of FRP from measurements of conductivity. Calculation of FRP is recommended for brevity and greater yield of data.