Water quality assessment is a critical aspect of environmental monitoring, requiring robust methodologies capable of handling uncertainties inherent in complex water quality data. This paper applies two innovative approaches for determining water quality: Fuzzy Parameterized Fuzzy Soft k-Neareat Neighbor (FPFS-kNN) and Picture Fuzzy Soft k-Nearest Neighbor (PFS-kNN). These methods leverage the unique characteristics of picture fuzzy soft matrices (pfs-matrices) and fuzzy parameterized fuzzy soft matrices (fpfs-matrices) to address the intricate nature of water quality parameters and associated uncertainties. The proposed methods are evaluated using a real-world water quality dataset, considering the accuracy, precision, recall, and F1-score performance metrics. Comparisons with traditional k-Nearest Neighbor (kNN) and other state-of-the-art methods demonstrate the superiority of FPFS-kNN and PFS-kNN in handling uncertainty and improving the accuracy of classification of water quality. This research contributes to advancing water quality assessment methodologies by introducing novel approaches that leverage the capabilities of pfs-matrices and fpfs-matrices.