Five hundred and seven British Friesian × Ayrshire bull calves were reared on a once-daily milk replacer feeding system from approximately 6 to 72 days of age in four naturally ventilated, unheated calf houses differing in design and degree of insulation. Environment within the calf houses was monitored each hour. Differences in design and degree of insulation produced small but significant differences in internal house temperature, humidity and air movement which had no measurable effect on calf daily live-weight gains. Calves receiving 0·6 kg/day compared with 0·3 or 0·4 kg/day of milk replacer plus concentrates ad libitum ate less concentrates but had higher daily live-weight gains (0·56 compared with 0·49 kg/day respectively) (P < 0·001). Rations supplying similar amounts of metabolizable energy from varying proportions of milk:concentrate resulted in comparable live-weight gains. Results suggest that calf viability was affected by level of nutrition and that low levels of milk replacer feeding resulted in higher mortality. Naturally ventilated, unheated calf houses were demonstrated to be suitable for rearing calves in south-west Scotland.