The Flow Cytometry Standard (FCS) for cytometric data (Dean et al.: Cytometry 11:323-332, 1990) provides for appending an ANALYSIS section to a data file, but it does not explicitly provide for recording the identities of individual cells. We propose an extension to the FCS definition in order to record the identity of each individual cell in list mode data. In order to do so, one first defines the subpopulations of cells to be identified and one than assigns a number to each defined population. For example, in an analysis in which peripheral blood mononuclear cells are labeled with antibodies to CD3, CD4, and CD8, the "negative" lymphocytes that are labeled with none of the antibodies could be identified as population 1, the CD3-CD4-CD8+ lymphocytes as population 2, etc. As the measured values from each cell are analyzed, the number that identifies the population to which that cell belongs is assigned as the value of an additional parameter. Since this procedure merely adds a new parameter, the only necessary extension to the FCS specification is that a particular name is recognized for this parameter. We propose that CS (abbreviation for cell subset) be recognized as the name and that CS be used as the value for the $PnN (parameter name) keyword in the FCS file TEXT. The CS values that are created can be used to aid in data analysis and can be permanently recorded in a conventional FCS file. A data file that is saved with CS values includes an explicit and integral record of the complete analysis, regardless of the complexity of the analysis.(ABSTRACT TRUNCATED AT 250 WORDS)