Patients with rheumatoid arthritis (RA) have significantly lower salivary and serum potassium (K) concentration, reduced total body K, and lower dietary K intake than healthy subjects. There may also be a subtle impairment in the hypothalamic-pituitary-adrenal (HPA) axis in patients with RA with both a poor cortisol secretion response as well as a lower adrenocorticotropin hormone (ACTH) response in relation to involved inflammatory factors. Patients with RA also exhibit an impaired Na+, K+-ATPase (NKA) activity which might promote the pro-inflammatory cytokine secretion seen in RA. I will use these facts to support the mechanism I propose. There are no qualitative differences between the effects of endogenous cortisol and exogenously applied synthetic glucocorticoids (GCs), which are widely used to treat RA. All effects are transmitted via the same receptor. The GC, cortisol, plays a role in normal K homeostasis and the reverse is also seen with higher K intake leading to higher cortisol secretion and biosynthesis. Results of a recent clinical trial showed elevated serum cortisol followed K supplementation. I suggest that this is what alleviated RA symptoms. I would like to suggest a "Cortisol-K" theory as a mechanism for De Coti-Marsh's proposed "K theory" while not precluding the possibility of eventual proof of a cure, possibly from effects of K inside cells other than the adrenal glands.