A good human-animal relation is important for pig welfare. To evaluate the effect of sows’ relation to human on their piglets’, we measured the influence of sow’s reaction to an unfamiliar human on their piglets’ reaction to the same human after weaning. We studied 24 sows and 96 of their piglets of the same breed. The sows were group housed during the gestating period and belonged to three different groups of 18 (±1) individuals. We observed their reaction to an unknown human approaching them in their pen, one by one, at 70 days of gestation. They were then classified in two categories whether they avoided human contact or not: shy (n=11) and docile (n=13). They were transferred to individual farrowing pens one week before farrowing. The piglets were weaned at 28 days and reared in groups of three from the same litter. Two male and two female piglets of each sow were chosen to be observed. At 32 or 33 days, each piglet was submitted to an individual human voluntary approach test (HVA) in a 2.7 m x 2.7 m room. After 5 minutes of isolation in this novel area, the same human as for the sows entered the room and sat still along a wall on a small stool. The behaviour of each piglet was recorded thanks to video analysis. A principal component analysis and a hierarchical clustering on principle components were performed to categorize the piglets according to their behaviours. Three categories were obtained that we named “shy” category (n=44), constituted of piglets staying away from human and freezing, “friendly” category (n=20), constituted of piglets coming and staying close to human, and “explorer” category (n=32), constituted of piglets not affected by the presence of the human and exploring the room. A Pearson's Chi-squared test was performed and showed a significant effect of the sow category on their piglets’ reaction to human (X=8.55, p=0.013). It appears that shy sows mostly produce shy piglets and that docile sows produce any of the three kind of piglets described. A shy sow had 59% chance to have shy piglets and only 9% chance to have friendly piglets. Those findings suggest that sows could transmit their reaction to humans to their piglets and that this is persistent event after weaning. If the transmission is accomplished non-genetically, taming shy sows could thus be favourable both for sow and piglet welfare.