The traditional air quality model initialization approach, in which one "cycles" the model, using one forecast to initialize the next, suffers from potential model drift, in which errors in one forecast simulation (possibly due to errors in the meteorological, emissions, or air quality simulations within the system) propagate into the next. For the MCNC/BAMS Numerical Air Quality Prediction System, we have attempted to ameliorate these errors by the use of a new algorithm that assimilates chemical observational data from monitor data provided by EPA's AIRNOW project. We have employed the algorithm in our summer-season daily MM5/SMOKE/MAQSIP-RT air quality forecasts. Here we present some caveats about the limitations of the algorithm, together with an analysis and interpretation of both the near-term forecast performance and of the persistence of the effects due to the perturbation introduced by the assimilation.