In antidiuresis, vasopressin (AVP) occupation of V2 receptors in renal collecting ducts activates adenylyl cyclase, resulting in increased intracellular cAMP levels, which activates protein kinase A (PKA). PKA phosphorylates both the cAMP responsive element binding protein, which induces aquaporin-2 (AQP2) transcription, and AQP2, which then is translocated to the apical membrane, allowing urine concentration. Lithium treatment often causes nephrogenic diabetes insipidus (NDI), which coincides with decreased AQP2 expression and which generally is ascribed to reduced adenylyl cyclase activity. However, the underlying mechanism by which lithium causes NDI is poorly understood. This study demonstrated that the mouse cortical collecting duct mpkCCD c14 cells are a good model; the deamino-8 D-arginine vasopressin (dDAVP)-induced endogenous AQP2 expression and plasma membrane localization was time-dependently reduced by treatment with clinically relevant lithium concentrations. Lithium did not affect AQP2 stability but decreased its mRNA levels. Surprising, the effect of lithium was cAMP independent; it did not alter AVP-stimulated cAMP production or PKA-dependent phosphorylation of AQP2 or cAMP responsive element binding protein. In vivo, kidney tissue of rats with lithium-induced NDI indeed generated less dDAVP-induced cAMP than that of controls, but this could be due to elevated blood AVP levels in rats with lithium-induced NDI. Indeed, Brattleboro rats, which lack endogenous AVP, with clamped blood dDAVP levels, showed no difference in dDAVP-generated cAMP generation between kidneys of rats with lithium-induced NDI and control rats. In conclusion, the first proper cell model to study lithium-induced NDI was developed, and it was demonstrated that the lithium-induced downregulation of AQP2 and development of NDI occur independent of adenylyl cyclase activity in vitro and in vivo.