Systematic error is a series of errors in accuracy that are consistent in a certain direction, while random errors are those which are caused by random and unpredictable variation in an experiment. Generally, systematic error is introduced by a problem that is consistent through an entire experiment. Random error is statistical fluctuations that are introduced by imprecision in measurement.
Systematic and random error are best contrasted by using examples. An example of random error would be weighing the same ring three times with the same scale and getting the different values of 17.1, 17.3 and 17.2 grams. Random errors tend to follow a normal distribution. An example of systematic error would be using an electric scale that reads 0.6 grams too high to take a series of masses. Every mass recorded would deviate from the true mass by 0.6 grams.
Both systematic and random error are types of experimental error, and minimizing them is key to a successful and meaningful experiment. Random error is generally corrected for by taking a series of repeated measurements and averaging them. Systematic error is more difficult to minimize because it is hard to detect. Using a second instrument to double-check readings is a good way to determine whether a certain instrument is introducing systematic error to a set of results.