The aim of this experiment was to study the effects of six different management routines at the time of farrowing on latency to first suckle, heat loss, weight gain and postnatal mortality. A total of 872 piglets from 67 loose housed sows in a commercial pig unit were subjected to one of six different management routines: control (CON n=14), no treatment; (CREEP n=13), placed in creep area; (UDDER n=10), placed at the udder; (DRY n=10), dried and placed back where found; (DRYCREEP n=9), dried and placed in creep area; and (DRYUDD n=11), dried and placed at the udder. The latency from birth to first suckle, rectal temperature at birth, 2 hours and 24 hours were measured for each piglet, in addition to weight at birth, 2 hours and 24 hours. Latency from birth to first suckle was shortest for piglets in the DRYUDD treatment, followed by the UDDER treatment. More live born piglets died in the UDDER treatment compared to the other treatments, but there were no other differences between the treatments with regards to postnatal mortality. There was a significant interaction between treatment and batch, with a significantly lower postnatal mortality in the DRYUDD treatment than CON in batch 2, but not in batch 1 and 3. Large litter sizes resulted in a higher postnatal mortality in all treatments, and tended to reduce latency to suckle. In conclusion, drying the piglets after birth and placing them at the udder resulted in reduced postnatal mortality in batch 2, but not in the other two batches. Despite having the largest mean litter size of the treatments, less than 10% of the piglets in DRYUDD died, which is remarkably low for loose housed sows. Regardless of treatment, several piglet-related factors were found to be highly important for postnatal mortality, such as the number of functional teats per piglet, birth weight, the latency from birth to first suckle, and rectal temperature at 2 hours after birth.