Nowadays, many web databases "hidden" behind their restrictive search interfaces (e.g., Amazon, eBay) contain rich and valuable information that is of significant interests to various third parties. Recent studies have demonstrated the possibility of estimating/tracking certain aggregate queries over dynamic hidden web databases. Nonetheless, tracking all possible aggregate query answers to report interesting findings (i.e., exceptions), while still adhering to the stringent query-count limitations enforced by many hidden web databases providers, is very challenging. In this paper, we develop a novel technique for tracking and discovering exceptions (in terms of sudden changes of aggregates) over dynamic hidden web databases. Extensive real-world experiments demonstrate the superiority of our proposed algorithms over baseline solutions.