A bioconcentration factor (BCF) database and a toxicokinetic model considering only biota-water partitioning and biotransformation were constructed for neutral organic chemicals in midge. The database contained quality-reviewed BCF and toxicokinetic data with variability constrained to within 0.5 to 1 log unit. Diverse conditions in exposure duration, flow setup , substrate presence, temperature, and taxonomic classification did not translate into substantial variability in BCF, uptake rate constant (k 1), or depuration rate constant (k T), and no systematic bias was observed in BCFs derived in unlabeled versus radiolabeled studies. Substance-specific biotransformation rate constants k M were derived by difference between the calculated biota-water partitioning coefficient (K BW) and experimental BCF for developing a midge biotransformation model. Experimental midge BCF was modeled as BCF = K BW /(1 + k M/ k 2) with log k M (k M in h-1) =-0.37 log K OW-0.06T (in K) + 18.87 (root mean square error [RMSE] = 0.60), log k 1 (k 1 in L kg wet.wt-1 h-1) =-0.0747 W (body weight in mg wet.wt) + 2.35 (RMSE = 0.48). The K BW value was estimated using midge biochemical composition and established polyparameter linear free energy relationships, and the diffusive elimination rate constant (k 2) was computed as k 2 = k 1 /K BW. The BCF model predicted >85% of BCFs that associated with neutral organic compounds (log K OW = 1.46-7.75) to within 1 log-unit error margin and had comparable accuracy similar to amphipod or fish models. A number of outliers and critical limitations of the k M model were identified and examined, and they largely reflected the inherent limitation of difference-derived k M , the lack of chemical diversity, and inadequate temperature variation in existing data. Future modeling efforts can benefit from more BCF and toxicokinetic observations of BCF on structurally diverse chemicals for model training, validation, and diagnosis.