Ketosis in dairy animals originates from negative energy status, associated increased absorption, and production of ketone bodies exceeding their use by the ruminants as an energy source. The present therapeutic experiment was carried out in 29 herds of Chilika buffaloes in 16 villages of three adjoining districts of Chilika Lake, Puri, Khurda, and Ganjam. Twenty Chilika buffaloes, detected positive for subclinical ketosis, were randomly selected for the study and divided into 2 groups, groups II and III, and were treated with hypertonic dextrose solution intravenously or gluconeogenic precursors along with nicotinamide orally, along with other supportive drugs in both the groups. Ten lactating Chilika buffaloes with no signs of ketosis and detected negative on Rothera test were included in the study as healthy controls (group I). Blood and milk samples were collected from all the 30 recruited buffaloes on days 0 (pre-treatment), 7, 14, and 28 for haematological and biochemical analysis. The subclinical ketosis in Chilika buffaloes did not have overt clinical signs. However, close examination revealed gradual drop in milk yield (100%), inappetence (59%), debility (46%), and uncoordinated gait (10%) without excitatory nervous signs. On day 7 following treatment, blood glucose concentration increased significantly. The mean serum triglyceride concentration of group III, treated with gluconeogenic precursors with nicotinamide, continued to decline significantly on subsequent observations. The serum enzyme activity, indicating status of liver function, declined following treatment in both the therapeutic groups. The intravenous administration of hypertonic dextrose solution compared to use of oral gluconeogenic precursors along with nicotinamide efficiently restored recovery from the subclinical ketosis in Chilika buffaloes.