The reductions in COD for both control and planted raceways are shown in Table 5. A significant level of COD reductions was noticed with all raceways and the planted raceways have performed better than the controls. The statistical analysis of the performance data (Table 8) confirms the removals as significant (p < 0.01). In experiment I, the rate of removal of COD with time, at the initial COD levels on day 1 (∼450mgl−1) was about 94%, as compared to that on day 5 (26.6mgl−1). The increase in removal efficiency becomes more evident with an increase in time (Table 5). This can be seen by comparing the residual COD remaining in the effluent at the end of each run in an experiment, to that of control (Fig. 6). For instance, in experiment I the ratio obtained by the COD present in the effluents coming out of the control RWs and that in the planted RWs, increased from 1 (on day 1) to approximately 2 (on day 5). Similar, but more encouraging results were obtained after the second dosage on day 10 where the ratio has increased again from 1 (on day 10) to nearly 7 (on day 15). Similarly, in experiment II, the ratio obtained for day 1 was 1 which has increased to 2.5 on day 5; subsequent to fresh dose of sewage on day 10 the ratio has increased from 1 to 12.4 on day 15 (Fig. 6). The ratio obtained (in experiment IV) by the percent of COD in the control, and in the presence of C.esculenta plants, increased from 1 (on day 1) to approximately 1.2 (on day 5) and from 1 (on day 15) to approximately 1.6 (on day 20). This indicated that the latter part (run 4) of the experiment IV was more efficient than the earlier part (run 1). A similar trend was observed in the rest of the experiments also. As mentioned earlier, the planted raceways have performed with better COD removals than the controls, for example, in experiment IV on day 5, COD level dropped from its initial 1650 to 213mgl−1 (87.1% removal) in the presence of plants. In controlled raceways, the removal was 84.2%.