The line of least squares gives an equation for a line that approximates the trend of the data set.  Using this equation will give what would be a reasonable expected output for a given input value.  By using the equation for the line of least squares, people can use previous trends in data to predict what might occur in the future.  One example of this implementation is predicting population.  Scientists use past data trends for rise in population over previous years to predict the possible population in ten or even one hundred years.

It could predict any number of growth criterion such as the growth charts used by doctors to predict a childs ultimate height, or the expected population in future years, or even the spread of disease.  The department of Transportation uses similar procedures to predict the number of accidents that are to be expected over holidays. 

The line of least squares is not always an accurate determination of an outcome.  Data can vary from the approximation.  Also, outliers can gravely influence the equation for the line of least squares which in turn can influence the accuracy of the predicted outcomes.  Also, a data set could appear to have a linear relation, when in actuality it is exponential.  This exponential relationship just may not appear in the small amount of data that has been collected.