The aim of the present study was to compare the effect of changing a range of biological traits on farm profit and greenhouse gas (GHG) emissions (expressed as carbon dioxide equivalent, CO2-eq.) in the UK dairy cow population. A Markov chain approach was used to describe the steady-state herd structure of the average milk-recorded UK dairy herd, as well as to estimate the CO2-eq. emissions per cow, and per kilogram of milk solids (MS). Effects of changing each herd production and fitness trait by one unit (e.g. 1 kg milk; 1% mastitis incidence) were assessed, with derived values for change in profit (economic values) being used in a multi-trait selection index. Of the traits studied, an increase in survival and reductions in milk volume, live weight, residual feed intake, somatic cell count, mastitis incidence, lameness incidence and calving interval were traits that would be both profitable and reduce CO2-eq. emissions per cow and per kg MS of a dairy herd. A multi-trait selection index was used to estimate the annual response in production and fitness traits and the economic response, with an estimate of annual profit per cow from selection on multiple traits. Milk volume, milk fat and protein yield, live weight, survival and dry matter intake were estimated to increase each year and body condition score, residual feed intake, somatic cell count, mastitis incidence, lameness incidence and calving interval were estimated to decrease, with selection on these traits estimated to result in an annual increase of 1% per year in GHG emissions per cow, but a reduction of 0·9% per unit product. Improved efficiencies of production associated with a reduction in milk volume (and increasing fat and protein content), live weight and feed intake (gross and metabolic efficiency, respectively), and increase in health, fertility and overall survival will increase farm annual profit of UK dairy systems and reduce their environmental impact.