A procedure for Stable Isotope Dilution Analysis in molecular Mass Spectrometry which does not require a methodological calibration graph and can be applied in combination with minimal labelling has been evaluated. This alternative approach is based on the determination of the molar fractions for each pure isotope pattern (natural abundance or labelled) contributing to the isotope pattern observed in the mixture of natural abundance and labelled molecules by multiple linear regression. Two labelled compounds, ¹³C₁-labelled or ¹³C₆-labelled phenol, were compared to study the influence of the number of ¹³C atoms in the labelled molecule. The procedure was evaluated by comparing the results obtained for the determination of phenol in NIST 1584 CRM by GC-EI-MS using the classical isotope dilution calibration procedure and the new procedure based on multiple linear regression of isotope patterns without a calibration graph. The results obtained using the proposed procedure agreed well with both the certified values and those obtained using the classical Isotope Dilution Mass Spectrometry (IDMS) calibration procedures. For the evaluated procedure, a full uncertainty budget determination has been developed taking into account all uncertainty sources, including those derived from the uncertainties in the isotope patterns of the natural and labelled compounds. The measurements with the ¹³C₁-labelled phenol provided lower propagated uncertainties in comparison to the use of ¹³C₆-labelled phenol.