ABSTRACT. Near-infrared/visible spectroscopy was used to develop predictive models for rapid measurement of five-day biochemical oxygen demand (BOD 5 ive-day biochemical oxygen demand (BOD 5 ) is a frequently measured index of organic loading in wastewater that requires five days of sample incubation to attain the test results. It is a measure of the impact that organic loading and subsequent microbial growth have on the dissolved oxygen (DO) concentration. BOD 5 is reported as mg/L DO depletion, and high levels of BOD 5 can adversely affect biological activity downstream from the discharge point.)Although BOD 5 is commonly used as a parameter in the determination of water quality, several problems have been identified: (1) the method fails to accurately characterize industrial toxic waste, which inhibits microorganism growth; (2) it can be difficult to obtain a good seed population of microbes to accurately predict BOD 5 for the environment from which the sample was taken; and (3) the BOD 5 test has a large number of variables, which are difficult to control, resulting in poor precision and reproducibility. These include the incubation temperature, seed quality, natural biological inhibitors, time taken to complete the test, and testing method accuracy.A portion of the oxygen consumption in BOD 5 tests can be due to nitrification. Unless it is inhibited, nitrification can cause a significant and typically undesired offset in the test results. This is especially of concern in samples with low-level BOD 5 . The nitrification is a lag phenomenon that is seen in unacclimated seed or for extreme temperatures of substrate (Swamee and Ojha, 1991). Usually allylthiourea or some other nitrification inhibitor is used to produce results that are more accurate and consistent. Nitrogenous oxygen demand accounts for 24% to 86% of the total BOD 5 (Koopman et al., 1989). From this, Koopman et al. estimated that 60% of violations of effluent discharge regulations are caused by nitrification in the BOD 5 test. A better test may remedy these problems.In addition, the 5-day lag between sampling and test results makes the data virtually useless for timely control decisions because the decisions would be based on 5-dayold data. Leighton and Euhus (1997) noted that process improvements are necessary for maintaining or reducing BOD while increasing food plant production. Online BOD 5 monitors would allow for more efficient processing of wastewater, and effluent could be more assuredly compliant with government regulations whether that wastewater is from urban treatment facilities, food processing plants, or other sources.Hassapis (1991a) was able to obtain two-minute BOD 5 values with a ±15% error via sample dilution into a whirl-bed reactor while controlling the dilution ratio to maintain constant oxygen consumption. The actual time of analysis varied between 2 and 10 minutes, depending on the reactor volume (Hassapis, 1991b). In the research, a simulation model was developed and experimentally verified. Based on his work in BOD 5 ...