Obesogenic diets and obesity are associated with cognitive impairment in both humans and animals. While the underlying mechanisms remain controversial, a key potential driver is increased inflammatory signalling associated with obesity. The anti-inflammatory antibiotic minocycline hydrochloride has been routinely used to depress microglial activity as it easily crosses the blood brain barrier. We used a rodent model to test the hypothesis that minocycline would alleviate the cognitive impairment produced by a high-fat high-sugar “cafeteria” diet. The study was a 2x2 design, where rats were exposed to either vehicle (syrup) or minocycline (40mg/kg/day) while consuming either regular chow or cafeteria diet. Memory was tested using novel object and place recognition tasks (NOR and NPR, respectively) and EchoMRI determined body composition across 6 weeks of intervention. Rats fed the cafeteria diet and vehicle were impaired on the hippocampal-dependent NPR at 2, 4 and 6 weeks but minocycline administration spared NPR performance in cafeteria-fed rats (similar exploration ratios as chow and vehicle). Of interest, chow rats treated with minocycline performed worse than those treated with vehicle. Cafeteria-fed rats irrespective of drug treatment consumed 150% more energy over the experiment and gained 100% more fat mass relative to rats fed chow. Faecal microbiota alpha diversity was reduced by both cafeteria diet and minocycline, but these reductions were no associated with performance on the NPR. However, abundances of specific OTUs within Bacteroides and Lactobacillus were associated with place task performance. Minocycline did not affect gene expression of IL-1B, TNF or IL6 in retroperitoneal white adipose tissue, although rats fed cafeteria diet and minocycline exhibited significantly increased expression of TLR-4. In summary, cafeteria diet produced persistent deficits in NPR that were prevented by minocycline cotreatment. The differences in behaviour observed correlated with differences in microbiome composition, but not with inflammatory gene expression in retroperitoneal fat.