This paper presents BBS (branch-and-bound skyline), a progressive algorithm for skyline queries that outperforms the existing NN (nearest neighbors) algorithm. Skyline queries identify points that are not dominated by any other point in all dimensions. NN, which uses divide-and-conquer and R-tree indexing, is efficient but has drawbacks such as high space overhead and duplicate elimination. BBS, based on nearest neighbor search, is IO optimal, accessing only nodes that may contain skyline points and avoiding duplicates. It is simpler to implement and performs better than NN in terms of CPU and IO costs, with significantly smaller space overhead. BBS is also efficient for various skyline query variations. The paper compares BBS with existing algorithms, showing that it outperforms NN by orders of magnitude in performance. BBS is optimal in terms of node accesses and is suitable for dynamic and constrained skyline queries. It is also efficient for ranked skyline queries, where the top-K points are returned based on a preference function. The algorithm is applicable to any dataset and dimensionality, using standard indexing structures. BBS is proven to be optimal in terms of node accesses and has lower memory overhead than NN. It is also efficient for constrained skyline queries, where the search space is restricted by constraints. The algorithm is suitable for dynamic skyline queries, where the dimensions are transformed based on user-defined functions. BBS is efficient for all types of skyline queries and is a promising solution for efficient skyline computation in databases.This paper presents BBS (branch-and-bound skyline), a progressive algorithm for skyline queries that outperforms the existing NN (nearest neighbors) algorithm. Skyline queries identify points that are not dominated by any other point in all dimensions. NN, which uses divide-and-conquer and R-tree indexing, is efficient but has drawbacks such as high space overhead and duplicate elimination. BBS, based on nearest neighbor search, is IO optimal, accessing only nodes that may contain skyline points and avoiding duplicates. It is simpler to implement and performs better than NN in terms of CPU and IO costs, with significantly smaller space overhead. BBS is also efficient for various skyline query variations. The paper compares BBS with existing algorithms, showing that it outperforms NN by orders of magnitude in performance. BBS is optimal in terms of node accesses and is suitable for dynamic and constrained skyline queries. It is also efficient for ranked skyline queries, where the top-K points are returned based on a preference function. The algorithm is applicable to any dataset and dimensionality, using standard indexing structures. BBS is proven to be optimal in terms of node accesses and has lower memory overhead than NN. It is also efficient for constrained skyline queries, where the search space is restricted by constraints. The algorithm is suitable for dynamic skyline queries, where the dimensions are transformed based on user-defined functions. BBS is efficient for all types of skyline queries and is a promising solution for efficient skyline computation in databases.