Background: Currently, there is no 'gold standard' for detecting patients with sensitivity to lactose. Biochemical investigation by a breath hydrogen test alone detects ,50% cases. Breath methane and symptoms are not recorded as standard practice. The clinical value of analysing C/T 13910 and G/A 22018 polymorphisms, strongly associated with lactose sensitivity, has not been established. Methods: Two hundred and ten patients with unexplained gut and systemic symptoms and controls were challenged with 50 g lactose. Breath hydrogen and methane were measured and symptoms recorded. All were genotyped for two polymorphisms, C/T 13910 and G/A 22018 . Results: CC 13910 /GG 22018 in 14.5%, CT 13910 /GA 22018 in 39% and TT 13910 /AA 22018 in 46.5%. One hundred percent of CC 13910 / GG 22018 were lactose sensitive having a breath hydrogen .20 ppm within 6 h and symptoms. But the breath hydrogen test lacked sensitivity and specificity in the other groups. There was elevated breath hydrogen in 21% of CT 13910 /GA 22018 and 15% of TT 13910 /AA 22018 by 6 h, whereas 17 and 30.9% had elevated breath methane alone. Breath methane and breath hydrogen with clinical symptoms improved sensitivity and specificity, increasing detection of lactose sensitivity in genotypes CT/GA and TT/AA from ,50 to .75%. Conclusions: The data presented define the current best practice for the clinical identification of lactose sensitivity. Patients were first genotyped. Those identified as CC with symptoms should immediately undertake a 12-week lactose-free diet. Those identified as CT or TT should undertake a breath hydrogen and methane test. Those positive for hydrogen or methane along with symptoms or with symptoms only, should also undertake a lactose-free diet. Those with high hydrogen without symptoms should be investigated for causes other than lactose sensitivity.