Abstract: The current study used multivariable logistic regression analysis to investigate associations between the intake frequencies of 13 food groups (or four diet groups) and infectious diseases. The analysis included 487849 participants from the UK Biobank, with 75209 participants diagnosed with infectious diseases. Participants reporting the highest intake frequency of processed meat (odds ratio OR = 1.0964, 95% confidence interval CI: 1.0622–1.1318) and red meat (OR = 1.0895, 95% CI: 1.0563–1.1239) had a higher risk of infectious diseases, compared with those with the lowest intake frequency. Consuming fish 2.0–2.9 times (OR = 0.8221, 95% CI: 0.7955–0.8496), cheese ≥ 5.0 times (OR = 0.882 2, 95% CI: 0.855 9–0.9092), fruit 3.0–3.9 servings (OR = 0.8867, 95% CI: 0.8661–0.9078), and vegetables 2.0–2.9 servings (OR = 0.9372, 95% CI: 0.9189–0.9559) per week were associated with a lower risk of infection. Low meat-eaters (OR = 0.9404, 95% CI: 0.9243–0.9567), fish-eaters (OR = 0.8391, 95% CI: 0.7887–0.8919), and vegetarians (OR = 0.9154, 95% CI: 0.8561–0.9778) had a lower risk of infectious diseases, compared with regular meat-eaters. The mediation analysis revealed that glycosylated hemoglobin, white blood cell count, and body mass index served as the mediators in the associations between diet and infectious diseases. The current study indicates that the intake frequency of food groups is a risk factor for infectious diseases, and fish-eaters have a lower risk of infection.