Endurance athletes fail to meet carbohydrate (CHO) guidelines for competition, which may be due to limited knowledge. However, the relationship between knowledge and practice in this population is unknown. To investigate this, we assessed the dietary intake in 50 athletes (37 females) who completed endurance events >= 2.5 h in duration and compared CHO intake against the carbohydrates for endurance athletes in competition questionnaire validated nutrition knowledge questionnaire, with specific questions related to CHO loading, pre-competition meal and during-competition intake. CHO-loading guidelines (10-12 g <middle dot> kg(-1) <middle dot> day(-1)) were met in practice by n = 5 (10%), but there was no relationship between identified requirements (range 0-12 g <middle dot> kg(-1) <middle dot> day(-1)) and actual intake (r(s) = 0.133, p = 0.358), with the n = 18 (36%) who correctly identified requirements, ingesting 6.1 +/- 1.9 g <middle dot> kg(-1) <middle dot> day(-1). CHO intake for pre-competition meal guidelines (1-4 g <middle dot> kg(-1)) was met in practice by n = 40 (80%), but there was no relationship between identified requirements (range 0 to >4 g <middle dot> kg(-1)) and actual intake (r(s) = 0.101, p = 0.487), with n = 19 (38%) who correctly identified guidelines requirements, ingesting 1.4 +/- 0.6 g <middle dot> kg(-1). CHO intake during-competition guidelines (60-90 g <middle dot> h(-1)) was met in practice by n = 18 (36%), but there was no relationship between the amounts of CHO required (range 30 to >90 g/h) and actual intake (r(s) = 0.028, p = 0.849), with n = 32 (64%) who correctly identified guidelines requirements, ingesting 56 +/- 20 g <middle dot> h(-1). Results show no relationship between the knowledge of CHO recommendations and practice, suggesting that theoretical knowledge does not guarantee the achievement of best practice and other important factors may ultimately determine practice.