Nursing in piglets relies on a particularly complex communication system, due in part to piglets being a precocial and a polytocous species lacking a milk cistern, such that milk is available at all teats simultaneously only for very short periods of time throughout the day. After lactation is established during the first couple of days, nursing becomes synchronous and cyclical, with all piglets suckling simultaneously every 30–70 min, over 20 times a day (Fraser, 1980; Lewis and Hurnik, 1985). These regular nursing bouts include five discrete phases that are observable in both domestic (Fraser, 1980) and wild breeds (Horrell, 1997). The first three phases involve the communication and coordination between the sow and her litter to ensure piglets arrive at and stimulate the udder to induce milk ejection (Algers, 1993; Bazer et al., 2001). Once the sow and piglets are synchronized and the udder has received adequate stimulation, the fourth phase, milk ejection, occurs (Fraser, 1980). The fifth and final phase of piglet nursing involves massage after milk is ejected. This period involves both massage and slow sucking (Fraser, 1980) and is very variable in length, depending on whether the sow or piglets terminate the nursing bout (Jensen et al., 1998) and on the age of the piglets (Horrell, 1997). The final massage is generally longer in duration than the massage prior to milk ejection (Algers and Jensen, 1985) and involves the expenditure of substantial amounts of energy (Klaver et al., 1981) that would seem to be counterproductive for animals trying to survive and grow. The objective of this experiment was to determine if there is a relationship between non-nutritive massage and growth in the young pig over the course of lactation. The suckling behaviour of 195 piglets from 20 litters was observed during at least three nursing bouts on each of three separate days. Piglets were weighed at birth and at 15 days post-partum, and 15-day growth rates were calculated. There was a significant difference between and within litters in suckling behaviour performed after milk ejection. Growth rates were negatively correlated with suckling during the final massage for individual piglets (r = -0.26; P < 0.003) and entire litters (r = -0.79; P < 0.001). Slower growing piglets also performed more suckling behaviour between nursing bouts r = -0.23; P < 0.05) and during incomplete nursing bouts (r = -0.30; P < 0.01) Piglets that performed the final massage consistently grew slower than those not involved with this costly behaviour. This data supports the hypothesis that massage after milk ejection is representative of piglets’ nutritional need. However, it appears that a lengthy massage does not result in an overall net return on piglet growth rate.