The cystic fibrosis transmembrane conductance regulator (CFTR) is a Cl^– channel gated by ATP-driven nucleotide-binding domain (NBD) dimerisation. To investigate the gating pathway of CFTR, we performed rate-equilibrium free energy relationships (REFER) measurements (1, 2). We used site-directed mutations as structural probes, and ATP and voltage as environmental probes of channel gating (1) and C1↔C2↔O scheme to describe CFTR channel gating (3). C1 represents the long duration closed state separating channel openings and C2↔O the bursting state. Transitions between C1↔C2 and C2↔O are described by the forward rate constants β1 and β2 and the backward rate constants α1 and α2, respectively. As CFTR is gated by intracellular ATP, C1↔C2 reflects agonist binding step, whereas C2↔O reflects the gating step (4). REFER analysis is applied to the gating step. β2 and α2 are used to generate a Brønsted plot (log (β2) plotted vs. log (β2/α2)). The slope of the line in a Brønsted plot (Φ) quantifies the relative extent to which the opening (β2) and closing (α2) rate constants change and provides an estimate of the temporal sequence of intermediate events during channel gating. Φ ranges between 0 and 1. When Φ is close to 1, the transition-state resembles an open-channel conformation and moves early during gating, whereas when Φ is close to 0, the transition-state resembles a closed-channel conformation and moves late during gating (1, 2). Brønsted plots of mutations (G550E, G551D, V562I, G551D-G550E, V562I-G550E) in the H5 α-helix of NBD1 (n = 3−14), different ATP concentrations (0.03−5 mM, n = 8−33) and a membrane voltage series (-100 − +100 mV, n = 5−7) yielded Φ values of 0.64 (r² = 0.91), 0.84 (r² = 0.55) and 0.18 (r² = 0.71), respectively. Good linear fits of the Brønsted plots of these perturbations indicate that CFTR is amenable to REFER analysis. We interpret our data to suggest that at the transition state, i) the structure of the H5 α-helix is a hybrid, which is more open than closed; ii) the ATP-binding sites are almost completely in the open state, arguing that the conformations of the ATP-binding sites change early in the CFTR gating pathway; iii) the CFTR pore is almost closed at the transition state, suggesting that the conformation of the CFTR pore changes late in the CFTR gating pathway. We conclude that there is a spatial gradient of Φ values from ~0.90 at the ATP-binding sites to ~0.20 at the CFTR pore. Thus, as proposed by Grosman et al (1), initiation at the effector site and propagation to the active site is likely to be a common theme for the gating of ion channels activated by ligands.