The study of the stars that explode as supernovae used to be a forensic study, working backwards from the remnants of the star. This changed in 1987 when the first progenitor star was identified in pre-explosion images. Currently, there are eight detected progenitors with another 21 non-detections, for which only a limit on the pre-explosion luminosity can be placed. This new avenue of supernova research has led to many interesting conclusions, most importantly that the progenitors of the most common supernovae, type IIP, are red supergiants, as theory has long predicted. However, no progenitors have been detected thus far for the hydrogen-free type Ib/c supernovae, which, given the expected progenitors, is an unlikely result. Also, observations have begun to show evidence that luminous blue variables, which are among the most massive stars, may directly explode as supernovae. These results contradict the current stellar evolution theory. This suggests that we may need to update our understanding.