Popular music artists record multiple albums that vary in quality over their careers. Folk psychology posits that music artists’ first albums are considered their best, whereas later albums draw fewer accolades; and that artists’ second albums are considered worse than their first—a phenomenon called the “sophomore slump.” The present work is the first large-scale multi-study attempt to test changes in album quality over time and whether a sophomore slump bias exists. Study 1 examined music critics, sampling all A, B, and C entries from The New Rolling Stone Record Guide (2,078 album reviews, 387 artists, 38 critics). Study 2 examined music fans, sampling crowdsourced Rate Your Music ratings of artists with a least one Rolling Stone top 500 album (4,030 album reviews, 254 artists). Using multilevel models—albums nested within artists (and artists nested within critics in Study 1)—both studies showed significant linear declines in ratings of artists’ album quality over artists’ careers; however, the linear effects were qualified by significantly positive quadratic effects, suggesting slightly convex patterns where declines were steeper among earlier (vs. later) albums. Controlling for these trends, a significant and substantial sophomore slump bias was observed for critics’ ratings, but not for fans’ ratings. Paradoxically, among critics, expertise—assessed via log number of albums reviewed—was inversely related to the strength of their sophomore slump bias. We discuss theoretical perspectives that may contribute to the observed effects, including regression to the mean, cognitive biases and heuristics, and social psychological accounts.