Current practice basing dialysis dose on urea distribution volume (V) has been questioned. We explored the impact on survival of scaling dialysis dose (Kt) to parameters reflective of metabolic activity. In a multicentre prospective cohort study of 1500 patients on thrice-weekly haemodialysis, body surface area (BSA) and resting energy expenditure (REE) were estimated using validated equations and physical activity by the Recent Physical Activity Questionnaire. Total energy expenditure (TEE) was estimated from REE and physical activity data. Kt was calculated from delivered (single-pool Kt/V)*Watson V. Kt/BSA, Kt/REE and Kt/TEE were then calculated at baseline and 6 monthly during follow-up for 2 years. In adjusted Cox models Kt/TEE, Kt/BSA, Kt/REE, in that order, had lower hazard ratios for death than single-pool Kt/V. On the basis of adjusted survival differences, putative minimum target doses were estimated for Kt/BSA as 27119 ml/m2 and Kt/TEE as 25.79 ml/kcal. We identified spKt/V values equivalent to these estimated targets, ranging from 1.4 to 1.8 in patient groups based on gender, body size and physical activity. For sedentary patients, the minimum target dose was 1.4 for large males, 1.5 for small males and 1.7 for women. For active patients the target was 1.8 irrespective of gender and body-weight. Patients achieving these individualised minimum targets had greater adjusted two-year survival compared to those achieving conventional minimum targets. Metabolic activity related parameters, such as Kt/TEE and Kt/BSA, may have a clinically important role in scaling haemodialysis dose. Using such parameters or their spKt/V equivalents to adjust minimum target doses based on gender, body size and habitual physical activity may have a positive impact on survival.