The backbone structure of a protein is largely determined by the / and w torsion angles. Thus, knowing these angles, even if approximately, will be very useful for protein-structure prediction. However, in a previous work, a sequence-based, real-value prediction of w angle could only achieve a mean absolute error of 548 (838, 358, 338 for coil, strand, and helix residues, respectively) between predicted and actual angles. Moreover, a real-value prediction of / angle is not yet available. This article employs a neural-network based approach to improve w prediction by taking advantage of angle periodicity and apply the new method to the prediction to / angles. The 10- fold-cross-validated mean absolute error for the new method is 388 (588, 338, 228 for coil, strand, and helix, respectively) for w and 258 (358, 228, 168 for coil, strand, and helix, respectively) for /. The accuracy of realvalue prediction is comparable to or more accurate than the predictions based on multistate classification of the /2w map. More accurate prediction of real-value angles will likely be useful for improving the accuracy of fold recognition and ab initio proteinstructure prediction.