Multiplying dietary NDF by particle size (PS) has been used as an estimate of physically effective fiber (peNDF). Our objectives were to (1) to compare the use of peNDF as dietary NDF × PS vs. use of individual NDF and PS descriptors in a physically adjusted NDF (paNDF) system when used with other factors to predict dry matter (DM) intake (DMI), rumination time, and ruminal pH in lactating dairy cows, and (2) leverage equations derived in a meta-analysis into an ensemble modeling system for predicting dietary physical and chemical characteristics required to maintain desired rumen conditions. Each response variable tested had 8 models in a 2 (peNDF, paNDF) × 2 (diet, diet+ruminal factors) × 2 (DM, as fed (AF) basis) factorial arrangement. PS descriptors were determined with the Penn State Particle Separator. Treatment means (n = 241) from 60 publications were used to derive models by backward elimination, weighted, mixed effect regression. Models containing peNDF terms had similar or lower prediction accuracy and precision than did models without peNDF terms. peNDF models of ruminal pH did not differ substantially from paNDF models. All variables from the meta-analyses were P < 0.05, and variables that entered the ensemble models included mean PS (MPS), AF or DM proportions retained on 19- and 8-mm sieves of the PSPS, DMI, diet concentrations of forage, forage NDF, crude protein, starch, NDF, and the interaction terms of starch × MPS, ADF/NDF, and rumination time/DMI. The peNDF system predicted that the minimum proportion of material (DM basis) retained on the 8-mm sieve should increase with decreasing forage NDF or dietary NDF. To maintain ruminal pH, an ensemble of peNDF models predicted that the minimum proportion of DM retained on the 8 mm sieve should increase with decreasing forage NDF or NDF; the minimum proportion of DM on the 8 mm sieve should increase as dietary starch increased, but also depends on the proportion of DM on the 19 mm sieve. Results of this study agree with and quantify known interrelationships between the chemical and physical forms of diets that affect ruminal pH in dairy cows, and offer the potential to generate feeding recommendations.