The objective of this paper is to examine the potential relationships between various tobacco blend components and Hoffmann analyte yields across three distinct lamina tobacco blend styles which are Virginia, Burley (uncased), and Oriental; additionally a 1:1 mixture of and Virginia/Burley was also included. For each blend style various cited Hoffmann blend precursors e.g. nitrogen, sugar, polyphenols, etc; and Hoffmann analyte yields were measured using a constant cigarette design, which has been described previously (Case et al, 2005). Correlations were found with these studies that indicate possible relationships between blend chemistry components and Hoffmann analytes in smoke, which, in the majority of cases agreed with published literature. Other analysis techniques included cluster analysis, by which precursors are grouped together; additionally linear regression analysis also identified blend components which allow prediction of certain Hoffmann analyte yields from tobacco precursors. The analysis techniques can also be used to illustrate that certain Hoffmann analyte yields can be correlated to more than one source of precursor. These relationships potentially indicated certain blend components which could, in future, be included with other cigarette design parameters in mathematically derived equations to predict Hoffmann analyte yields in mainstream smoke. However the correlations obtained should not be interpreted as 'cause and effect', in that a specific blend component is definitely a precursor of a particular Hoffmann analyte.