The paper presents improved thermodynamic parameters for predicting RNA duplex stability. The parameters, derived from enthalpy and free-energy changes for helix formation by 45 RNA oligonucleotide duplexes, include one parameter for duplex initiation and 10 parameters for helix propagation. These parameters predict melting temperatures of most oligonucleotide duplexes within 5°C, which is considered a good performance for the nearest-neighbor model. The sequences were chosen to maximize the reliability of secondary structure predictions and minimize experimental errors in ΔG° at 37°C. The parameters also account for free-energy changes at dangling ends, terminal mismatches, and internal G-U mismatches, as well as helix initiation at hairpin loops, internal loops, or internal bulges. The predictive capability of the parameters is tested against experimental data, showing average deviations of about 6%, 8%, and 2°C for ΔG°T, ΔH°, and t_m, respectively. The parameters are useful for predicting RNA secondary structures, with 82% of the major stems in 142 randomly chosen tRNA sequences predicted within 2 base pairs using a specific algorithm.The paper presents improved thermodynamic parameters for predicting RNA duplex stability. The parameters, derived from enthalpy and free-energy changes for helix formation by 45 RNA oligonucleotide duplexes, include one parameter for duplex initiation and 10 parameters for helix propagation. These parameters predict melting temperatures of most oligonucleotide duplexes within 5°C, which is considered a good performance for the nearest-neighbor model. The sequences were chosen to maximize the reliability of secondary structure predictions and minimize experimental errors in ΔG° at 37°C. The parameters also account for free-energy changes at dangling ends, terminal mismatches, and internal G-U mismatches, as well as helix initiation at hairpin loops, internal loops, or internal bulges. The predictive capability of the parameters is tested against experimental data, showing average deviations of about 6%, 8%, and 2°C for ΔG°T, ΔH°, and t_m, respectively. The parameters are useful for predicting RNA secondary structures, with 82% of the major stems in 142 randomly chosen tRNA sequences predicted within 2 base pairs using a specific algorithm.