Under nominally CO₂/bicarbonate-free conditions (Hepes-buffer), proton transmission between cardiomyocytes occurs by means of protonated mobile buffers passively permeating gap junctions. Ventricular myocytes isolated from humanely killed guinea-pigs were AM-loaded with the pH-sensitive dye, carboxy-SNARF-1, and superfused with Hepes-buffered Tyrode solution (37°C) while imaging pH_i confocally. It has been shown (Swietach & Vaughan-Jones, 2004) that the apparent junctional permeability constant (P_H^app) can be calculated from the longitudinal pH_i gradient that is established when a cell-pair is partially exposed to a weak acid or weak base. Using a double-barrelled micropipette, end-to-end cell- pairs were dually microperfused with two microstreams, one of which contained 20-30mM ammonium or 80-120mM acetate. 30μM cariporide was included in both microstreams to block sarcolemmal Na⁺-H⁺ exchange. Exposure of the distal end of one cell produced a large pH_i gradient in the cell-pair that was maintained for the duration of dual microperfusion. P_H^app was estimated from the discontinuity of the longitudinal pH_i profile that coincided with the junctional region. To examine the pH_i-dependence of P_H^app, the pH_i of the cell-pair was adjusted to different values (by prepulsing uniformly with extracellular ammonium or acetate) before applying the dual microstream. P_H^app was maximal (5.17(±0.23)×10^-4 cm/s, N=14) at pH 7.0 and declined significantly (P<0.05) at both higher and lower pH_i (2.20(±0.38)×10^-4 cm/s at pH 7.36, N=13 and 1.72(±0.48)×10^-4 cm/s at pH 6.33, N=8). In the presence of reduced extracellular Ca²⁺ (from 2mM to 0.5mM) and increased intracellular Ca²⁺ buffering (by preloading cells with 100μM BAPTA-AM), the maximum P_H^app did not change significantly (4.60(±0.49)×10^-4 cm/s at pH 6.92, N=17), and P_H^app was again reduced at both higher and lower pH_i (2.16(±0.59)×10^-4 cm/s at pH 7.30, N=5 and 2.87(±0.43)×10^-4 cm/s at pH 6.39, N=10). The reduction of P_H^app with acidosis under these conditions was, however, significantly smaller than the reduction under normal conditions (P<0.05), suggesting that during intracellular acidosis, gap junctions may be closed by a combination of raised [H⁺]_i and [Ca²⁺]_i. Gating in the alkaline direction, on the other hand, appears to be Ca²⁺_i-independent. Such gating will influence the local spread of acid or base within the myocardium.