Past studies have repeatedly demonstrated the technical feasibility to produce polyhydroxyalkanoate (PHA) using bacterial biomass of mixed microbial cultures (MMCs). Commercial quality grades of poly(3-hydroxybutyrate-co-3-hydroxyvalerate), PHBV, can be produced with control of average monomer composition. However, demonstration of PHBV production and recovery with quality control of molecular weight (MW) distribution has been lacking in the research literature. Towards this goal, a workflow has been developed for characterizing molecular weight control by thermal treatment pre-processing of dried PHA-rich biomass before solvent extraction. Dimethyl carbonate (DMC) was a suitable solvent in this workflow in the routine evaluation of extractable PHA. From assessments of DMC extraction using differential scanning calorimetry, 125 °C was selected for nominally 100 percent extraction yield independent of polymer 3-hydroxyvalerate (3HV) content (2 to 41 wt.% 3HV) and molecular weight (100 to 1400 kDa). Intrinsic viscosity measurements of PHBV in DMC at 60 °C was used for molecular weight monitoring. Mark-Houwink constants, α (0.738 ± 0.010) and LogK (-2.016 ± 0.025), were estimated for a PHBV co-polymer blend having 36 wt.% 3HV. A model of random scission supported that weight average molecular weight (M_w) was a more robust metric, compared to number average molecular weight (M_n), for assessing the polymer scission rates. During isothermal heat treatment for a given biomass batch, interpreted scission rate was reproducible and commonly, but not always, constant in time. Scission rates between biomass batches were also variable. Measured properties of the polymer in the biomass (thermal stability, biomass PHA content, PHBV grade, initial moisture content) could not be correlated to this observed batch-to-batch variation of scission rate. Molecular weight loss before extraction did not influence the melting temperatures of the co-polymer blends of PHBV evaluated over a wide sub-eutectic range of average 3HV content. Molecular weight changes for these PHBV co-polymer blends were considered to have likely influenced the nature of blend 3HV distribution, and consequently, crystallization behaviour. Molecular weight loss effects on crystallization behaviour at constant PHBV average 3HV wt.% content could then have contributed to the observed variability for glass transition temperatures and melting enthalpies. However, a reproducible correlation between this variability and MW change was not observed.