All-in-all-out (AIAO) systems are becoming popular in the industry. This is leading to the realization of the costs of slower growing pigs. Focus is now shifting towards total herd growth average and range in that growth. Statistically, variation can be measured by standard deviation (SD) and coefficient of variation (CV), both of which make use of the “bell curve”. The bell curve shows how many individuals are clustered around the average and how far they stray from the average (“deviate”). A narrow bell is desirable. Mean (average), median (middle observation in a range), mode (most frequent observation), minimum and maximum observations, SD, and CV are all useful in measuring variability. Measuring variation on-farm is difficult to do and the proper amount of pigs must be weighed to obtain acceptable results. The industry has developed management techniques to deal with variability such as cross fostering. Variation can affect production via sort losses, barn utilization, and effective phase feeding. Barn space is effected because faster growing pigs will go to market quicker and slower growing pigs will take several weeks longer. This will offset the flow of production. The most obvious cost of variation is in the carcass grading. In 2002, producers lost about $6.9 million based on Mitchell’s grading grid due to variability! Causes of this variation can be narrowed down to genetic and environmental. The goal of producers should be to minimize the influence of the environment on variability so that genetics is the lone contributor to variation. This description of variation can also be called “intrinsic” (factors based on the actual pig) or “extrinsic” (based on external factors such as the environment). Based on social behaviour, if average performance is good and variation low, it can be assumed that conditions are good because of little aggression. When performance drops it can be assumed that there is competition for resources such as food and water. This results in restriction of feed to some pigs, which will likely increase the variability. When variation in a facility is relatively low, management is the best key to handling it. If variation is high then it should be looked at to reduce variability. Reasonable CV targets are 20% weaning weights, 12 to 15% nursery exit weights and 8 to 12% for weight at first pull. If the CV is above target, look at reducing variability. If it is on target, work on managing the variability.