Discussed are two sources of error in the decrepitation method of study of liquid inclusions. Certain minerals were found to contain inclusions that leaked prior to breaking, with the result that the decrepitation temperatures recorded for these minerals exceeded the visually determined temperatures of filling by as much as 150 degrees Centigrade. Samples of other minerals studied contained secondary inclusions which formed at lower temperatures than primary inclusions in the same sample. The secondary inclusions in these samples began to decrepitate at lower temperatures than the primary inclusions and continued to decrepitate as heating continued, with the result that the start of decrepitation of the primary inclusions was masked.Neither of the sources of error can be detected by decrepitation tests alone. Decrepitation tests, therefore, must be preceded by optical examinations both at room temperature and with the heating stage. The optical examinations would have to be done in such detail as to eliminate the need for decrepitation tests.Discussion is included which indicates the necessity for consideration of these sources of error in the interpretation of the results of any study of liquid inclusions.