Potassium channels share a common selectivity filter and their diversity is given by how they are gated open. TASK-2 (K2P5.1) and TASK-3 (K2P9.1) are two-pore-domain K⁺ channels gated by extracellular pH (pH_o), which play important roles in renal bicarbonate reabsorption (Warth et al. 2004) and excitability (Brickley et al. 2007) respectively. Whilst TASK-3, a TASK subfamily member, is inhibited by lowering pH_o, the TALK subfamily member TASK-2 is activated by alkalinization. The molecular determinants of pH_o-gating for these channels have been identified. TASK-3 senses pH_o through pore-mouth residue H98 (N103 in TASK-2) (Kim et al. 2000; Rajan et al. 2000). Sensing of pH_o in TASK-2 is believed to take place at R224 (V221 in TASK-3) located in the outermost portion of helix TM4 near the channel pore (Niemeyer et al. 2007). The unusual pK_a of 8.0 of the arginine sensor in TASK-2 has been attributed to its hydrophobic environment. Mutation of sensor residues of TASK-2 and -3 to neutral amino acids abolishes pH_o sensitivity. Here we report patch-clamp experiments using the recombinant channels to analyse the result of pH_o-sensor transplantation between TASK-2 and TASK-3 further to validate their identification. The position equivalent to TASK-2 R224 is occupied by V221 in TASK-3. Replacing this residue by a titratable histidine (to facilitate protonation-deprotonation at near-physiological pH) in the background of the pH_o-insensitive TASK-3-H98N led to a recovery of pH_o-sensitivity. The observed pK_a values were 6.0±0.1 (mean ± SEM, n= 8) for the WT TASK-3 and 6.9±0.1 (n=7) for the double mutant TASK-3-H98N-V221H. The Hill coefficient n_H was decreased from a value of 2.1±0.1 to 0.65±0.1, near that for TASK-2, in the double mutant. The K⁺/Na⁺ selectivity was also decreased in TASK-3-H98N-V221H, suggestive of an electrostatic effect leading to some form of pore collapse induced by the presence of the positively charged residue at position 221 in TASK-3. The position equivalent to TASK-3 H98 is occupied by N103 in TASK-2. The mutation of N103 for H, to create a TASK-3 sensor in the background of pH_o-insensitive TASK-2-R224A also restored H⁺-gating in this mutant. The observed pK_a values were 8.0 ± 0.1 (n= 6) for WT TASK-2 and 7.6 ± 0.1 (n=7) for the double mutant TASK-2-R224A-N103H. Our results support the idea that R224 acting with a markedly shifted pK_a is the pH_o sensor of TASK-2. The protonation-deprotonation of R224 would have a direct effect on the electrostatic potential of the selectivity filter and would regulate open probability at the pore by modulating K⁺ occupancy.