Height-biomass regressions were used to produce non-destructive aboveground biomass assessments for Spartina foliosa. We performed a statistical analysis of these regression models using data collected in Punta Banda Estuary, Baja California, Mexico. We found these models to be inconsistent with available data, on the basis of lack of homoscedasticity and randomness of residuals. The allometric scaling of aboveground plant biomass and corresponding leaf area was tested as a basis for alternative biomass assessment. We show that regression models using leaf area result in simpler, more reliable non-destructive estimations of Spartina foliosa biomass.