We reanalyze the catalogs of molecular clouds in the Local Group to determine the parameters of their mass distributions in a uniform manner .
The analysis uses the error-in-variables method of parameter estimation which accounts not only for the variance of the sample when drawn from a parent distribution but also for errors in the mass measurements .
Testing the method shows that it recovers the underlying properties of cumulative mass distribution without bias while accurately reflecting uncertainties in the parameters .
Clouds in the inner disk of the Milky Way follow a truncated power-law distribution with index \gamma = -1.5 \pm 0.1 and maximum mass of 10 ^ { 6.5 } ~ { } M _ { \odot } .
The distributions of cloud mass for the outer Milky Way and M33 show significantly steeper indices ( \gamma _ { \mathrm { OMW } } = -2.1 \pm 0.2 and \gamma _ { \mathrm { M 33 } } = -2.9 \pm 0.4 ) with no evidence of a cutoff .
The mass distribution of clouds in the Large Magellanic Cloud has a marginally steeper distribution than the inner disk of the Milky Way ( \gamma = -1.7 \pm 0.2 ) and also shows evidence of a truncation with a maximum mass of 10 ^ { 6.5 } ~ { } M _ { \odot } .
The mass distributions of molecular clouds vary dramatically across the Local Group , even after accounting for the systematic errors that arise in comparing heterogeneous data and catalogs .
These differences should be accounted for in studies that aim to reproduce the molecular cloud mass distributions or in studies that use the mass spectrum as a parameter in a model .